Skip to main content

Resultados relacionados

Resultados totales (Incluyendo duplicados): 63
Encontrada(s) 1 página(s)

e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Contribución de congreso (ConferenceOutput)

Detection of People Boarding/Alighting a Metropolitan Train using Computer Vision

  • Belloc, M.
  • Velastin Carroza, Sergio Alejandro
  • Fernández, R.
  • Jara, M.
This paper has been presented at : 9th International Conference on Pattern Recognition Systems (ICPRS 2018), Pedestrian detection and tracking have seen a major progress in the last two decades. Nevertheless there are always appli-cation areas which either require further improvement or that have not been sufficiently explored or where production level performance (accuracy and computing efficiency) has not been demonstrated. One such area is that of pedestrian monitoring and counting in metropolitan railways platforms. In this paper we first present a new partly annotated dataset of a full-size laboratory observation of people boarding and alighting from a public transport vehicle. We then present baseline results for automatic detection of such passengers, based on computer vi-sion, that could open the way to compute variables of interest to traffic engineers and vehicle designers such as counts and flows and how they are related to vehicle and platform layout., The authors gratefully acknowledge the Chilean National Science and Technology Council (Conicyt) for its funding under grants CONICYT-Fondecyt Regular nos. 1140209 (“OBSERVE”) , 1120219, and 1080381 . S.A. Velastin is grateful to funding received from the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development and demonstration under grant agreement no. 600371, el Ministerio de Economía y Competitividad (COFUND2013-51509) and Banco Santander. Finally, we are grateful to NVIDIA for its donation as part of its academic GPU Grant Program.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Instructor activity recognition through deep spatiotemporal features and feedforward Extreme Learning Machines

  • Nida, Nudrat
  • Yousaf, Muhammad Haroon
  • Irtaza, Aun
  • Velastin Carroza, Sergio Alejandro
Human action recognition has the potential to predict the activities of an instructor within the lecture room. Evaluation of lecture delivery can help teachers analyze shortcomings and plan lectures more effectively. However, manual or peer evaluation is time-consuming, tedious and sometimes it is difficult to remember all the details of the lecture. Therefore, automation of lecture delivery evaluation significantly improves teaching style. In this paper, we propose a feedforward learning model for instructor's activity recognition in the lecture room. The proposed scheme represents a video sequence in the form of a single frame to capture the motion profile of the instructor by observing the spatiotemporal relation within the video frames. First, we segment the instructor silhouettes from input videos using graph-cut segmentation and generate a motion profile. These motion profiles are centered by obtaining the largest connected components and normalized. Then, these motion profiles are represented in the form of feature maps by a deep convolutional neural network. Then, an extreme learning machine (ELM) classifier is trained over the obtained feature representations to recognize eight different activities of the instructor within the classroom. For the evaluation of the proposed method, we created an instructor activity video (IAVID-1) dataset and compared our method against different state-of-the-art activity recognition methods. Furthermore, two standard datasets, MuHAVI and IXMAS, were also considered for the evaluation of the proposed scheme., We gratefully acknowledge the support of NVIDIA Corporation with the donation of the Titan X Pascal GPU used for research work carried in Centre for Computer Vision Research (C2VR) at University of Engineering and Technology Taxila, Pakistan. Sergio A Velastin acknowledges funding by the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for Research, Technological Development and Demonstration under grant agreement no. 600371, el Ministerio de Economía y Competitividad (COFUND2013-51509), and Banco Santander.We are also very thankful to participants, faculty, and postgraduate students of Computer Engineering Department who took part in the data acquisition phase.Without their consent, this work was not possible.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Contribución de congreso (ConferenceOutput)

DA-VLAD: Discriminative action vector of locally aggregated descriptors for action recognition

  • Murtaza, Fiza
  • Yousaf, Muhammad Haroon
  • Velastin Carroza, Sergio Alejandro
This paper has been presented at : 25th IEEE International Conference on Image Processing (ICIP 2018), In this paper, we propose a novel encoding method for the representation of human action videos, that we call Discriminative Action Vector of Locally Aggregated Descriptors (DA-VLAD). DA-VLAD is motivated by the fact that there are many unnecessary and overlapping frames that cause non-discriminative codewords during the training process. DA-VLAD deals with this issue by extracting class-specific clusters and learning the discriminative power of these codewords in the form of informative weights. We use these discriminative action weights with standard VLAD encoding as a contribution of each codeword. DA-VLAD reduces the inter-class similarity efficiently by diminishing the effect of common codewords among multiple action classes during the encoding process. We present the effectiveness of DA-VLAD on two challenging action recognition datasets: UCF101 and HMDB51, improving the state-of-the-art with accuracies of 95.1% and 80.1% respectively., We gratefully acknowledge the support of NVIDIA Corporation with the donation of the Titan X Pascal GPU used for this research. We also acknowledge the support from the Directorate of Advance Studies, Research and Technological development (ASR) & TD, University of Engineering and Technology Taxila, Pakistan. Sergio A Velastin acknowledges funding by the Universidad Carlos III de Madrid, the European Unions Seventh Framework Programme for research, technological development and demonstration under grant agreement n 600371, el Ministerio de Economia y Competitividad (COFUND2013-51509) and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

On the economics of stand-alone renewable hybrid power plants in remote regions

  • Petrakopoulou, Foteini Konstantina
In recent years ever more examples of regions that have managed to achieve or orientate themselves toward renewable energy sufficiency are emerging. However, actions to create energy autonomy are mainly the result of isolated activities and they are less driven from fully organized movements. In addition, total energy independence without the support of a centralized electrical grid is yet to be achieved. The objectives of this work are to investigate the associated costs of stand-alone renewable hybrid power plants on a Greek island and compare them to the cost of the currently used fossil-fuel-based conventional plant. The plants examined here are designed to fully cover the electricity needs of the island. Islands may face numerous energy problems and rely heavily on foreign and environmentally-harmful fuels. It is shown that the relatively high cost of electricity of such a remote region can increase the competitiveness and promote the wider incorporation of technologies based on renewable energy sources that may, in other cases, seem economically inferior to business-as-usual energy solutions., Fontina Petrakopoulou would like to thank the Universidad Carlos III de Madrid, the European Union's Seventh Framework Programme for research, technological development and demonstration (grant agreements no 600371 and 332028), the Ministerio de Economía y Competitividad (COFUND2014-51509) the Ministerio de Educación, cultura y Deporte (CEI-15-17) and Banco Santander
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Monomolecular and Bimolecular Recombination of Electron-Hole Pairs at the Interface of a Bilayer Organic Solar Cell

  • Hahn, Tobias
  • Tscheuschner, Steffen
  • Kahle, Frank-Julian
  • Reichenberger, Markus
  • Athanasopoulos, Stavros
  • Saller, Christina
  • Bazan, Gillermo C.
  • Nguyen, Thuc-Quyen
  • Strohriegl, Peter
  • Bässler, Heinz
  • Köhler, Anna
While it has been argued that field-dependent geminate pair recombination (GR) is important, this process is often disregarded when analyzing the recombination kinetics in bulk heterojunction organic solar cells (OSCs). To differentiate between the contributions of GR and nongeminate recombination (NGR) the authors study bilayer OSCs using either a PCDTBTtype polymer layer with a thickness from 14 to 66 nm or a 60 nm thick p-DTS(FBTTh2)(2) layer as donor material and C-60 as acceptor. The authors measure JV-characteristics as a function of intensity and charge-extraction-by-linearly-increasing-voltage-type hole mobilities. The experiments have been complemented by Monte Carlo simulations. The authors find that fill factor (FF) decreases with increasing donor layer thickness (L-p) even at the lowest light intensities where geminate recombination dominates. The authors interpret this in terms of thickness dependent back diffusion of holes toward their siblings at the donor-acceptor interface that are already beyond the Langevin capture sphere rather than to charge accumulation at the donor-acceptor interface. This effect is absent in the p-DTS(FBTTh2)(2) diode in which the hole mobility is by two orders of magnitude higher. At higher light intensities, NGR occurs as evidenced by the evolution of s-shape of the JV-curves and the concomitant additional decrease of the FF with increasing layer thickness., The authors acknowledge financial support by the Bavarian State Ministry of Science, Research, and the Arts through the Collaborative Research Network “Solar Technologies go Hybrid”, by the Volkswagen foundation and by the German Science Foundation DFG through the doctoral training center “GRK 1640.” This project further received funding from the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development and demonstration under grant agreement no. 600371, el Ministerio de Economía y Competitividad (COFUND2014-51509), el Ministerio de Educación, cultura y Deporte (CEI-15-17), and Banco Santander. M.R. additionally acknowledges support from the Hanns Seidel Foundation for a stipend through funds from the German Ministry of Education and Research (BMBF). T.-Q.N. thanks the Office of Naval Research (#N000141410076) for the support. Furthermore, the authors would like to thank the anonymous referees for helpful suggestions.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Dynamic Spatio-Temporal Bag of Expressions (D-STBoE) Model for Human Action Recognition

  • Nazir, Saima
  • Yousaf, Muhammad Haroon
  • Nebel, Jean-Cristophe
  • Velastin Carroza, Sergio Alejandro
This article belongs to the Section Intelligent Sensors, Human action recognition (HAR) has emerged as a core research domain for video understanding and analysis, thus attracting many researchers. Although significant results have been achieved in simple scenarios, HAR is still a challenging task due to issues associated with view independence, occlusion and inter-class variation observed in realistic scenarios. In previous research efforts, the classical bag of visual words approach along with its variations has been widely used. In this paper, we propose a Dynamic Spatio-Temporal Bag of Expressions (D-STBoE) model for human action recognition without compromising the strengths of the classical bag of visual words approach. Expressions are formed based on the density of a spatio-temporal cube of a visual word. To handle inter-class variation, we use class-specific visual word representation for visual expression generation. In contrast to the Bag of Expressions (BoE) model, the formation of visual expressions is based on the density of spatio-temporal cubes built around each visual word, as constructing neighborhoods with a fixed number of neighbors could include non-relevant information making a visual expression less discriminative in scenarios with occlusion and changing viewpoints. Thus, the proposed approach makes the model more robust to occlusion and changing viewpoint challenges present in realistic scenarios. Furthermore, we train a multi-class Support Vector Machine (SVM) for classifying bag of expressions into action classes. Comprehensive experiments on four publicly available datasets: KTH, UCF Sports, UCF11 and UCF50 show that the proposed model outperforms existing state-of-the-art human action recognition methods in term of accuracy to 99.21%, 98.60%, 96.94 and 94.10%, respectively., Sergio A. Velastin is grateful for funding received from the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development and demonstration under grant agreement N° 600371, el Ministerio de Economía, Industria y Competitividad (COFUND2013-51509) el Ministerio de Educación, Cultura y Deporte (CEI-15-17) and Banco Santander. Muhammad Haroon Yousaf received funding from the Higher Education Commission, Pakistan for Swarm Robotics Lab under the National Centre for Robotics and Automation (NCRA). The authors also acknowledge support from the Directorate of ASR&TD, University of Engineering and Technology Taxila, Pakistan.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

TAB: Temporally aggregated bag-of-discriminant-words for temporal action proposals

  • Murtaza, Fiza
  • Yousaf, Muhammad Haroon
  • Velastin Carroza, Sergio Alejandro
In this work, we propose a new method to generate temporal action proposals from long untrimmed videos named Temporally Aggregated Bag-of-Discriminant-Words (TAB). TAB is based on the observation that there are many overlapping frames in action and background temporal regions of untrimmed videos, which cause difficulties in segmenting actions from non-action regions. TAB solves this issue by extracting class-specific codewords from the action and background videos and extracting the discriminative weights of these codewords based on their ability to discriminate between these two classes. We integrate these discriminative weights with Bag of Word encoding, which we then call Bag-of-Discriminant-Words (BoDW). We sample the untrimmed videos into non-overlapping snippets and temporally aggregate the BoDW representation of multiple snippets into action proposals using a binary classifier trained on trimmed videos in a single pass. TAB can be used with different types of features, including those computed by deep networks. We present the effectiveness of the TAB proposal extraction method on two challenging temporal action detection datasets: MSR-II and Thumos14, where it improves upon state-of-the-art with recall rates of 87.0% and 82.0% respectively at a temporal intersection over union ratio of 0.8., We gratefully acknowledge the support of NVIDIA Corporation with the donation of the TITAN Xp GPU used for this research. We also acknowledge the support from the Directorate of Advance Studies, Research and Technological development (ASR & TD), University of Engineering and Technology Taxila, Pakistan. Sergio A. Velastin acknowledges funding by the Universidad Carlos III de Madrid, the European Unions Seventh Framework Programme for research, technological development and demonstration under grant agreement no. 600371, el Ministerio de Economía y Competitividad (COFUND2013-51509) and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Contribución de congreso (ConferenceOutput)

Vehicle Detection Using Alex Net and Faster R-CNN Deep Learning Models: A Comparative Study

  • Espinosa, Jorge E.
  • Velastin Carroza, Sergio Alejandro
  • Branch, John W.
This paper has been presented at : 5th International Visual Informatics Conference (IVIC 2017), This paper presents a comparative study of two deep learning models used here for vehicle detection. Alex Net and Faster R-CNN are compared with the analysis of an urban video sequence. Several tests were carried to evaluate the quality of detections, failure rates and times employed to complete the detection task. The results allow to obtain important conclusions regarding the architectures and strategies used for implementing such network for the task of video detection, encouraging future research in this topic., S.A. Velastin is grateful to funding received from the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development and demonstration under grant agreement no. 600371, el Ministerio de Economía y Competitividad (COFUND2013-51509) and Banco Santander. The authors wish to thank Dr. Fei Yin for the code for metrics employed for evaluations. Finally, we gratefully acknowledge the support of NVIDIA Corporation with the donation of the GPUs used for this research. The data and code used for this work is available upon request from the authors.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Tape casting manufacturing of thick Li4Ti5O12 ceramic electrodes with high areal capacity for lithium-ion batteries

  • Torre Gamarra, Carmen de la
  • Sotomayor Lozano, María Eugenia
  • Bucheli Erazo, Wilmer Oswaldo
  • Amarilla, José Manuel
  • Sánchez, Jean Yves
  • Levenfeld Laredo, Belén
  • Varez, Alejandro
In this work we report the preparation of additive-free ceramic Li4Ti5O12(LTO) electrodes with high thickness(∼250mum) by the conventional tape casting method followed by binder removal and sintering. By optimizingthe sintering process, high porosity (25-35 %), needed for a good electrolyte wetting, was achieved. In addition,the use of an Ar/H2reducing atmosphere during the thermal treatments increases the required electronic conductivity of the sintered electrode due to the combination of a carbon coating coming from incomplete binder decomposition and the partially reduction of Ti4+. Due to the absence of binders, the final electrode has almost 100 wt. % active material, leading to very high volumetric (250 mA h cm-3) and areal (6-8 mA h cm-2) capacities. This process paves the way for large-scale production of thick electrodes with applications in all-solid-state batteries (ASSBs), leading to more safety lithium-ion batteries (LIBs)., We thank Agencia Estatal de Investigación (AEI)/Fondo Europeo de Desarrollo Regional (FEDER/UE) and Comunidad de Madrid for funding the projects MAT2016-78632-C4-3-R. RTI2018-095425-B-I00 and MATERYENER3CM S2013/MIT-2753 program. Authors also thank the ICTS-Centro Nacional de Microscopía Electrónica (UCM, Madrid) for instrumental facilities and Dr. Esteban Urones for fruitful discussion about carbon coating. J.-Y. Sanchez acknowledges the CONEX Programme, funding received from Universidad Carlos III de Madrid, the European Union's Seventh Framework Programme for research, technological development and demonstration (Grant agreement no. 600371), Spanish Ministry of Economy and Competitiveness (COFUND2013-40258) and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Concurrent enhancement of percolation and synchronization in adaptive networks

  • Eom, Young-Ho
  • Boccaletti, Stefano
  • Caldarelli, Guido
Co-evolutionary adaptive mechanisms are not only ubiquitous in nature, but also beneficial for the functioning of a variety of systems. We here consider an adaptive network of oscillators with a stochastic, fitness-based, rule of connectivity, and show that it self-organizes from fragmented and incoherent states to connected and synchronized ones. The synchronization and percolation are associated to abrupt transitions, and they are concurrently (and significantly) enhanced as compared to the non-adaptive case. Finally we provide evidence that only partial adaptation is sufficient to determine these enhancements. Our study, therefore, indicates that inclusion of simple adaptive mechanisms can efficiently describe some emergent features of networked systems' collective behaviors, and suggests also self-organized ways to control synchronization and percolation in natural and social systems., Y.-H.E. and G.C. acknowledge FET Project MULTIPLEX (nr. 317532), FET Project SIMPOL (nr. 610704) and FET Project DOLFINS (nr. 640772). Y.-H.E. would like to thank the Universidad Carlos III de Madrid, the European Union's Seventh Framework Programme for research, technological development and demonstration under grant agreement (no. 600371), el Ministerio de Economía y competitividad (COFUND2014-51509) and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Contribución de congreso (ConferenceOutput)

Silhouette-based human action recognition with a multi-class support vector machine

  • González, Luis
  • Velastin Carroza, Sergio Alejandro
  • Acuña Leiva, Gonzalo
This paper has been presented at : 9th International Conference on Pattern Recognition Systems (ICPRS 2018), Computer vision systems have become increasingly popular, being used to solve a wide range of problems. In this paper, a computer vision algorithm with a support vector machine (SVM) classifier is presented. The work focuses on the recognition of human actions through computer vision, using a multi-camera dataset of human actions called MuHAVi. The algorithm uses a method to extract features, based on silhouettes. The challenge is that in MuHAVi these silhouettes are noisy and in many cases include shadows. As there are many actions that need to be recognised, we take a multiclass classification ap-proach that combines binary SVM classifiers. The results are compared with previous results on the same dataset and show a significant improvement, especially for recognising actions on a different view, obtaining overall accuracy of 85.5% and of 93.5% for leave-one-camera-out and leave-one-actor-out tests respectively., Sergio A Velastin has received funding from the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development and demonstration under grant agreement no. 600371, el Ministerio de Economía, Industria y Competitividad (COFUND2013-51509) el Ministerio de Educación, cultura y Deporte (CEI-15-17) and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Economic and Environmental Considerations for Zero-emission Transport and Thermal Energy Generation on an Energy Autonomous Island

  • Petrakopoulou, Foteini Konstantina
The high cost and environmental impact of fossil-fuel energy generation in remote regions can make renewable energy applications more competitive than business-as-usual scenarios. Furthermore, energy and transport are two of the main sectors that significantly contribute to global greenhouse gas emissions. This paper focuses on the generation of thermal energy and the transport sector of a fossil fuel-based energy independent island in Greece. We evaluate (1) technologies for fully renewable thermal energy generation using building-specific solar thermal systems and (2) the replacement of the vehicle fleet of the island with electric and hydrogen-fueled vehicles. The analysis, based on economic and environmental criteria, shows that although solar thermal decreases greenhouse gases by 83%, when compared to the current diesel-based situation, it only becomes economically attractive with subsidy scenarios equal to or higher than 50%. However, in the transport sector, the sum of fuel and maintenance costs of fuel-cell and electric vehicles is found to be 45% lower than that of the current fleet, due to their approximately seven times lower fuel cost. Lastly, it will take approximately six years of use of the new vehicles to balance out the emissions of their manufacturing phase., Fontina Petrakopoulou would like to thank the Universidad Carlos III de Madrid, the European Union's Seventh Framework Programme for research, technological development and demonstration (grant agreements no 600371 and 332028), the Ministerio de Economía y Competitividad (COFUND2014-51509) the Ministerio de Educación, cultura y Deporte (CEI-15-17) and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

End-to-End Temporal Action Detection using Bag of Discriminant Snippets (BoDS)

  • Murtaza, Fiza
  • Yousaf, Muhammad Haroon
  • Velastin Carroza, Sergio Alejandro
  • Qian, Yu
Detecting human actions in long untrimmed videosis a challenging problem. Existing temporal action detectionmethods have difficulties in finding the precise starting andending time of the actions in untrimmed videos. In this letter, wepropose a temporal action detection framework based on a Bagof Discriminant Snippets (BoDS) that can detect multiple actionsin an end-to-end manner. BoDS is based on the observationthat multiple actions and the background classes have similarsnippets, which cause incorrect classification of action regionsand imprecise boundaries. We solve this issue by finding the keysnippetsfrom the training data of each class and compute theirdiscriminative power which is used in BoDS encoding. Duringtesting of an untrimmed video, we find the BoDS representationfor multiple candidate proposals and find their class label basedon a majority voting scheme. We test BoDS on the Thumos14 andActivityNet datasets and obtain state-of-the-art results. For thesports subset of ActivityNet dataset, we obtain a mean AveragePrecision (mAP) value of 29% at 0.7 temporal intersection overunion (tIoU) threshold. For the Thumos14 dataset, we obtain asignificant gain in terms of mAP i.e., improving from 20.8% to31.6% at tIoU=0.7., This work was supported by the ASR&TD, University of Engineering and Technology (UET) Taxila, Pakistan. The work of S. A. Velastin was supported by the Universidad Carlos III de Madrid, the European Unions Seventh Framework Program for research, technological development, and demonstration under Grant 600371, el Ministerio de Economia y Competitividad (COFUND2013-51509), and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Detecting, tracking and counting people getting on/off a metropolitan train using a standard video camera

  • Velastin Carroza, Sergio Alejandro
  • Fernández, Rodrigo
  • Espinosa, Jorge E.
  • Bay, Alessandro
The main source of delays in public transport systems (buses, trams, metros, railways) takes place in their stations. For example, a public transport vehicle can travel at 60 km per hour between stations, but its commercial speed (average en-route speed, including any intermediate delay) does not reach more than half of that value. Therefore, the problem that public transport operators must solve is how to reduce the delay in stations. From the perspective of transport engineering, there are several ways to approach this issue, from the design of infrastructure and vehicles to passenger traffic management. The tools normally available to traffic engineers are analytical models, microscopic traffic simulation, and, ultimately, real-scale laboratory experiments. In any case, the data that are required are number of passengers that get on and off from the vehicles, as well as the number of passengers waiting on platforms. Traditionally, such data has been collected manually by field counts or through videos that are then processed by hand. On the other hand, public transport networks, specially metropolitan railways, have an extensive monitoring infrastructure based on standard video cameras. Traditionally, these are observed manually or with very basic signal processing support, so there is significant scope for improving data capture and for automating the analysis of site usage, safety, and surveillance. This article shows a way of collecting and analyzing the data needed to feed both traffic models and analyze laboratory experimentation, exploiting recent intelligent sensing approaches. The paper presents a new public video dataset gathered using real-scale laboratory recordings. Part of this dataset has been annotated by hand, marking up head locations to provide a ground-truth on which to train and evaluate deep learning detection and tracking algorithms. Tracking outputs are then used to count people getting on and off, achieving a mean accuracy of 92% with less than 0.15% standard deviation on 322 mostly unseen dataset video sequences., Sergio A. Velastin is grateful for funding received from the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development and demonstration under grant agreement N 600371, el Ministerio de Economía, Industria y Competitividad (COFUND2013-51509) el Ministerio de Educación, Cultura y Deporte (CEI-15-17) and Banco Santander. Rodrigo Fernandez and Sergio A. Velastin gratefully acknowledge the Chilean National Science and Technology Council (Conicyt) for its funding under CONICYT-Fondecyt Regular Grant Nos. 1120219, 1080381 and 1140209 (“OBSERVE”).
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Ultra-thick battery electrodes for high gravimetric and volumetric energy density Li-ion batteries

  • Sotomayor Lozano, María Eugenia
  • Torre Gamarra, Carmen de la
  • Levenfeld Laredo, Belén
  • Sánchez, Jean Yves
  • Varez, Alejandro
  • Kim, Guk Tae
  • Varzi, Alberto
  • Passerini, Stefano
Lithium-ion batteries, already capable of addressing the requirements of electric vehicles and renewable electricity storage, will be in the focus for the next decade. However, the transition from the small batteries for portable electronic to ultra-high capacity batteries requires safety improvements and cost-cutting efforts. Intensive research has been performed in the last three decades to optimize specific and volumetric capacity of electrodes. Recently the attention is focusing on increasing the electrodes areal capacity to enable the substantial reduction of the current collectors, porous separator, and electrolyte resulting in large gravimetric and volumetric energy density improvements as well as cost savings. Here we propose a lithium-ion battery based on thick, additive-free ceramic negative and positive electrodes, i.e. LTO (Li4Ti5O12) and LFP (LiFePO4), obtained by a solvent-free technology called powder extrusion moulding. Against all odds, the LTO/LFP cell based on these thick ceramic electrodes (areal capacity of 13.3 inA h cm(-2)) achieved the very high energy of about 23.9 mW h cm(-2). Such an excellent performance, achieved simulating a typical day-night charge/discharge cycle, opens the way for the deployment of low cost, effective household (and small industry) energy storage in combination with photovoltaic energy harvesting., Authors would like to thank financial support received from Spanish Government MICINN (MAT2016-78362-C4-3-R project), and Madrid regional Government (MATERYENER3CM S2013/MIT-2753 program). They also thank the ICTS-Centro Nacional de Microscopia Electronica (UCM, Madrid) for TEM support. J.-Y. Sanchez acknowledges the CONEX Programme, funding received from Universidad Carlos III de Madrid, the European Union's Seventh Framework Programme for research, technological development and demonstration (Grant agreement n. 600371), Spanish Ministry of Economy and Competitiveness (COFUND2013-40258) and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Brain hemodynamic activity during viewing and re-viewing of comedy movies explained by experienced humor

  • Jaaskelainen, Liro P.
  • Pajula, Juha
  • Tohka, Jussi
  • Lee, Hsin-Ju
  • Kuo, Wen-Jui
  • Lin, Fa-Hsuan
Humor is crucial in human social interactions. To study the underlying neural processes, three comedy clips were shown twice to 20 volunteers during functional magnetic resonance imaging (fMRI). Inter-subject similarities in humor ratings, obtained immediately after fMRI, explained inter-subject correlation of hemodynamic activity in right frontal pole and in a number of other brain regions. General linear model analysis also indicated activity in right frontal pole, as well as in additional cortical areas and subcortically in striatum, explained by humorousness. The association of the right frontal pole with experienced humorousness is a novel finding, which might be related to humor unfolding over longer time scales in the movie clips. Specifically, frontal pole has been shown to exhibit longer temporal receptive windows than, e.g., sensory areas, which might have enabled processing of humor in the clips based on holding information and reinterpreting that in light of new information several (even tens of) seconds later. As another novel finding, medial and lateral prefrontal areas, frontal pole, posterior-inferior temporal areas, posterior parietal areas, posterior cingulate, striatal structures and amygdala showed reduced activity upon re-viewing of the clips, suggesting involvement in processing of humor related to novelty of the comedic events., This study was supported by the Academy of Finland (project numbers 138145 and 276643 to IPJ, 130275 and 263785 to JT, FiDiPro to FHL), aivoAALTO project funding of the Aalto University, Ministry of Science and Technology, Taiwan (MOST 104-2314-B-002-238, MOST 103-2628-B-002-002-MY3, MOST 104-2410-H-010 -003 -MY2), National Health Research Institute, Taiwan (NHRI-EX104-10247EI) and Ministry of Economic Affairs, Taiwan (100-EC-17-A-19-S1-175). JT has received funding from the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development and demonstration under grant agreement no. 600371, el Ministerio de Economía y Competitividad (COFUND2013-40258) and Banco Santander. We thank Taiwan Mind & Brain Imaging Center (TMBIC) and National Cheng-Chi University for instrument availability for this work.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

High mass loading additive-free LiFePO[subíndice 4] cathodes with 500 μm thickness for high areal capacity Li-ion batteries

  • Torre Gamarra, Carmen de la
  • Sotomayor Lozano, María Eugenia
  • Sánchez, Jean Yves
  • Levenfeld Laredo, Belén
  • Varez, Alejandro
  • Laïk, Barbara
  • Pereira Ramos, Jean-Pierre
We report the preparation of thick ceramic electrodes of the olivine LiFePO4 (LFP) with high mass loading. These electrodes are preparated by means of Powder Extrusion Moulding (PEM), which is a technology easily scalable and cheap. These LFP cathodes are additive-free (neither binder nor extra carbon black) with ~0.5 mm thickness, allowing to develop very high areal capacity (13.7 mA h cm−2). By means of a strict control of sintering process, the carbon coating of the commercial LFP powder remains and the decomposition of the active material is prevented. The optimized self-supported LFP cathode presents good cyclability over 20 cycles at C/10 with no capacity loss. The good electrochemical performance of these novel LFP thick electrodes and their non-flammability make them interesting candidates for both mobile and stationary applications., Authors would like to thank financial support received from Spanish Government MICINN (MAT2016-78362-C4-3-R project), and Madrid regional Government (MATERYENER3CM S2013/MIT-2753 program). They also thank the ICTS-Centro Nacional de Microscopía Electrónica (UCM, Madrid) for instrumental facilities and Dr. Esteban Urones for fruitful discussion about carbon coating. J.-Y. Sanchez acknowledges the CONEX Programme, funding received from Universidad Carlos III de Madrid, the European Union's Seventh Framework Programme for research, technological development and demonstration (Grant agreement n. 600371), Spanish Ministry of Economy and Competitiveness (COFUND2013-40258) and Banco Santander. C. de la Torre-Gamarra would like to thank Universidad Carlos III de Madrid for the mobility grant received for her predoctoral stay at Institut de Chimie et des Matériaux Paris Est (Thiais – France)., Publicado
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Exploiting deep residual networks for human action recognition from skeletal data

  • Huy Hieu, Pham
  • Khoudour, Louahdi
  • Crouzil, Alain
  • Zegers, Pablo
  • Velastin Carroza, Sergio Alejandro
The computer vision community is currently focusing on solving action recognition problems in real videos, which contain thousands of samples with many challenges. In this process, Deep Convolutional Neural Networks (D-CNNs) have played a significant role in advancing the state-of-the-art in various vision-based action recognition systems. Recently, the introduction of residual connections in conjunction with a more traditional CNN model in a single architecture called Residual Network (ResNet) has shown impressive performance and great potential for image recognition tasks. In this paper, we investigate and apply deep ResNets for human action recognition using skeletal data provided by depth sensors. Firstly, the 3D coordinates of the human body joints carried in skeleton sequences are transformed into image-based representations and stored as RGB images. These color images are able to capture the spatial-temporal evolutions of 3D motions from skeleton sequences and can be efficiently learned by D-CNNs. We then propose a novel deep learning architecture based on ResNets to learn features from obtained color-based representations and classify them into action classes. The proposed method is evaluated on three challenging benchmark datasets including MSR Action 3D, KARD, and NTU-RGB + D datasets. Experimental results demonstrate that our method achieves state-of-the-art performance for all these benchmarks whilst requiring less computation resource. In particular, the proposed method surpasses previous approaches by a significant margin of 3.4% on MSR Action 3D dataset, 0.67% on KARD dataset, and 2.5% on NTU-RGB +D dataset., This research was carried out at the Cerema Research Center (CEREMA) and Toulouse Institute of Computer Science Research (IRIT), France. Sergio A Velastin is grateful for funding received from the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development and demonstration under grant agreement n 600371, el Ministerio de Economía, Industria y Competitividad (COFUND2013-51509) el Ministerio de Educación, Cultura y Deporte (CEI-15-17) and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Learning to Recognize 3D Human Action from A New Skeleton-based Representation Using Deep Convolutional Neural Networks

  • Pham, Huy-Hieu
  • Khoudour, Louahdi
  • Crouzil, Alain
  • Zegers, Pablo
  • Velastin Carroza, Sergio Alejandro
Recognizing human actions in untrimmed videos is an important challenging task. An effective 3D motion representation and a powerful learning model are two key factors influencing recognition performance. In this paper we introduce a new skeletonbased representation for 3D action recognition in videos. The key idea of the proposed representation is to transform 3D joint coordinates of the human body carried in skeleton sequences into RGB images via a color encoding process. By normalizing the 3D joint coordinates and dividing each skeleton frame into five parts, where the joints are concatenated according to the order of their physical connections, the color-coded representation is able to represent spatio-temporal evolutions of complex 3D motions, independently of the length of each sequence. We then design and train different Deep Convolutional Neural Networks (D-CNNs) based on the Residual Network architecture (ResNet) on the obtained image-based representations to learn 3D motion features and classify them into classes. Our method is evaluated on two widely used action recognition benchmarks: MSR Action3D and NTU-RGB+D, a very large-scale dataset for 3D human action recognition. The experimental results demonstrate that the proposed method outperforms previous state-of-the-art approaches whilst requiring less computation for training and prediction., This research was carried out at the Cerema Research Center (CEREMA) and Toulouse Institute of Computer Science Research (IRIT), Toulouse, France. Sergio A. Velastin is grateful for funding received from the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for Research, Technological Development and demonstration under grant agreement N. 600371, el Ministerio de Economia, Industria y Competitividad (COFUND2013-51509) el Ministerio de Educación, cultura y Deporte (CEI-15-17) and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Spatio-Temporal Image Representation of 3D Skeletal Movements for View-Invariant Action Recognition with Deep Convolutional Neural Networks

  • Hieu Pham, Huy
  • Salmane, Houssam
  • Khoudour, Louahdi
  • Crouzil, Alain
  • Zegers, Pablo
  • Velastin Carroza, Sergio Alejandro
This article belongs to the Special Issue Deep Learning-Based Image Sensors, Designing motion representations for 3D human action recognition from skeleton sequences is an important yet challenging task. An effective representation should be robust to noise, invariant to viewpoint changes and result in a good performance with low-computational demand. Two main challenges in this task include how to efficiently represent spatio-temporal patterns of skeletal movements and how to learn their discriminative features for classification tasks. This paper presents a novel skeleton-based representation and a deep learning framework for 3D action recognition using RGB-D sensors. We propose to build an action map called SPMF (Skeleton Posture-Motion Feature), which is a compact image representation built from skeleton poses and their motions. An Adaptive Histogram Equalization (AHE) algorithm is then applied on the SPMF to enhance their local patterns and form an enhanced action map, namely Enhanced-SPMF. For learning and classification tasks, we exploit Deep Convolutional Neural Networks based on the DenseNet architecture to learn directly an end-to-end mapping between input skeleton sequences and their action labels via the Enhanced-SPMFs. The proposed method is evaluated on four challenging benchmark datasets, including both individual actions, interactions, multiview and large-scale datasets. The experimental results demonstrate that the proposed method outperforms previous state-of-the-art approaches on all benchmark tasks, whilst requiring low computational time for training and inference., This research was carried out at the Cerema Research Center and Informatics Research Institute of Toulouse, Paul Sabatier University, France. The authors would like to express our thanks to all the people who have made helpful comments and suggestions on a previous draft. S.A. Velastin is grateful to funding received from the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development and demonstration under grant agreement no. 600371, el Ministerio de Economía y Competitividad (COFUND2013-51509) and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Detection of Motorcycles in Urban Traffic Using Video Analysis: A Review

  • Espinosa, Jorge E.
  • Velastin Carroza, Sergio Alejandro
  • Branch, John W.
Motorcycles are Vulnerable Road Users (VRU) and as such, in addition to bicycles and pedestrians, they are the traffic actors most affected by accidents in urban areas. Automatic video processing for urban surveillance cameras has the potential to effectively detect and track these road users. The present review focuses on algorithms used for detection and tracking of motorcycles, using the surveillance infrastructure provided by CCTV cameras. Given the importance of results achieved by Deep Learning theory in the field of computer vision, the use of such techniques for detection and tracking of motorcycles is also reviewed. The paper ends by describing the performance measures generally used, publicly available datasets (introducing the Urban Motorbike Dataset (UMD) with quantitative evaluation results for different detectors), discussing the challenges ahead and presenting a set of conclusions with proposed future work in this evolving area., This work was supported in part by the Departamento Administrativo de Ciencia, Tecnología e Innovación (COLCIENCIAS) Project: Reduccion de Emisiones Vehiculares Mediante el Modelado y Gestion Optima de Trafico en Areas Metropolitanas, Caso Medellín, Area Metropolitana del Valle de Aburra, under Grant 111874558167 and Grant CT 049-2017, in part by the Universidad Nacional de Colombia under Project HERMES 25374, and in part by NVIDIA Corporation for the donation of GPUs. The work of Sergio A. Velastín was supported in part by the Universidad Carlos III de Madrid, in part by the European Union's Seventh Framework Programme for Research, Technological Development and Demonstration under Grant 600371, in part by the El Ministerio de Economía y Competitividad under Grant COFUND2013-51509, and in part by the Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Contribución de congreso (ConferenceOutput)

Feature Similarity and Frequency-Based Weighted Visual Words Codebook Learning Scheme for Human Action Recognition

  • Nazir, Saima
  • Yousaf, Muhammad Haroon
  • Velastin Carroza, Sergio Alejandro
This paper has been presented at : 8th Pacific-Rim Symposium, PSIVT 2017., Human action recognition has become a popular field for computer vision researchers in the recent decade. This paper presents a human action recognition scheme based on a textual information concept inspired by document retrieval systems. Videos are represented using a commonly used local feature representation. In addition, we formulate a new weighted class specific dictionary learning scheme to reflect the importance of visual words for a particular action class. Weighted class specific dictionary learning enriches the scheme to learn a sparse representation for a particular action class. To evaluate our scheme on realistic and complex scenarios, we have tested it on UCF Sports and UCF11 benchmark datasets. This paper reports experimental results that outperform recent state-of-the-art methods for the UCF Sports and the UCF11 dataset i.e. 98.93% and 93.88% in terms of average accuracy respectively. To the best of our knowledge, this contribution is first to apply a weighted class specific dictionary learning method on realistic human action recognition datasets., Sergio A Velastin acknowledges funding by the Universidad Carlos III de Madrid, the European Unions Seventh Framework Programme for research, technological development and demonstration under grant agreement n 600371, el Ministerio de Economía y Competitividad (COFUND2013-51509) and Banco Santander. Authors also acknowledges support from the Directorate of ASR and TD, University of Engineering and Technology Taxila, Pakistan.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Contribución de congreso (ConferenceOutput)

Motorcycle Classification in Urban Scenarios using Convolutional Neural Networks for Feature Extraction

  • Espinosa, Jorge E.
  • Velastin Carroza, Sergio Alejandro
  • Branch, John W.
This paper has been presented at 8th International Conference of Pattern Recognition Systems., This paper presents a motorcycle classification system for urban scenarios using Convolutional Neural Network (CNN). Significant results on image classification has been achieved using CNNs at the expense of a high computational cost for training with thousands or even millions of examples. Nevertheless, features can be extracted from CNNs already trained. In this work AlexNet, included in the framework CaffeNet, is used to extract features from frames taken on a real urban scenario. The extracted features from the CNN are used to train a support vector machine (SVM) classifier to discriminate motorcycles from other road users. The obtained results show a mean accuracy of 99.40% and 99.29% on a classification task of three and five classes respectively. Further experiments are performed on a validation set of images showing a satisfactory classification., S.A. Velastin is grateful to funding received from the Universidad Carlos III de Madrid, the European Union's Seventh Framework Programme for research, technological development and demonstration under grant agreement no. 600371, el Ministerio de Economía y Competitividad (COFUND2013-51509) and Banco Santander
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Mechanisms of social movement success: conversation, displacement and disruption, Mecanismos para entender los logros de los movimientos: conversación, desplazamiento y disrupción

  • Fishman, Robert Michael
  • Everson, David W.
A great deal of social movement scholarship tends to assume - and in some cases explicitly argues - that disruption is the primary mechanism through which protest movements win major concessions from the holders of power. Nonetheless, some studies and much empirical evidence provide a strong basis to argue that other paths to social movement success also exist. The importance of discourse and framing has also been highlighted in a number of studies but we argue that the full contribution of talk itself to movement success is best captured through the concept of "conversation" and an examination of the preconditions for its viability. The successful displacement of power-holders by protest movements, although a less common pathway to success than disruption and conversation, also deserves conceptual and empirical attention. In this paper we make this set of distinctions conceptually explicit, differentiating analytically between these three mechanisms of movement success and delineating the conditions required for each mechanism to prove viable and 'useful'. We rely on extensive examples drawn from movements and protest events in the United States, Spain and Portugal, using this empirical material, as well as the existing theoretical literature, as our basis for constructing a conceptual argument on the ideal typical distinction between these three mechanisms and the conditions that allow them to operate. We also take up the questions of whether, and when, movement actors can successfully combine these mechanisms or - alternatively - find themselves pressed to pursue one or another of these pathways to success in a relatively 'pure' form, Buena parte de la investigación en movimientos sociales tiende a asumir - y en algunos casos argumenta de manera explícita - que la disrupción es el principal mecanismo a través del cual los movimientos de protesta obtienen importantes concesiones por parte de quienes detentan el poder. No obstante, algunos estudios y mucha evidencia empírica proporcionan una base sólida para sostener la existencia de otras vías para el logro de los movimientos sociales. Si bien la importancia del discurso y el enmarcado se ha puesto de manifiesto en una serie de estudios, nosotros sostenemos que el significado completo que la palabra tiene para el éxito de los movimientos se capta mejor a partir del concepto de "conversación" y un examen de sus condiciones de viabilidad. La sustitución o el desplazamiento de la titularidad del poder, a pesar de ser una vía menos frecuente para el logro de los movimientos de protesta, también merece atención conceptual y empírica. En este artículo hacemos conceptualmente explícitas este conjunto de distinciones, diferenciando de manera analítica entre estos tres mecanismos de éxito de un movimiento y delineando las condiciones requeridas por cada uno de ellos para demostrarse viable y "útil". Nos basamos en extensos ejemplos extraídos de movimientos y eventos de protesta en los Estados Unidos, España y Portugal, utilizando este material empírico así como la literatura teórica existente como base para construir nuestro argumento conceptual sobre la distinción de estos tres mecanismos como tipos ideales y las condiciones que les permiten ser operativos. También planteamos la cuestión de hasta qué punto - y cuándo - los actores de movimientos pueden combinar con éxito estos mecanismos o verse forzados a seguir una u otra de estas vías hacia el logro de una forma relativamente "pura", For extremely useful feedback on an earlier version, we wish to thank participants in the 2015 Midterm Conference of the ESA RN25 (European Sociological Association's research network on social movements) on, "From Contention to Social Change: Rethinking the Consequences of Social Movements and Cycles of Protest" as well as participants in the SPAM workshop at the University of Notre Dame and both the special issue editors and the anonymous readers. Also, Fishman gratefully acknowledges funding support received through the CONEX program from the Universidad Carlos III de Madrid, the European Union's Seventh Framework Program for research, technological development and demonstration under grant agreement number 600371, Spain's Ministerio de Economía y Competitividad (COFUND2013-40258), and the Banco de Santander. Fishman also acknowledges support by the Spanish Ministry of Economy and Competitiveness (grant number CSO2013-41035-P).
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Contribución de congreso (ConferenceOutput)

Multi-view Human Action Recognition using Histograms of Oriented Gradients (HOG) Description of Motion History Images (MHIs)

  • Murtaza, Fiza
  • Yousaf, Muhammad Haroon
  • Velastin Carroza, Sergio Alejandro
This paper has been presented at : 13th International Conference on Frontiers of Information Technology (FIT), In this paper, a silhouette-based view-independent human action recognition scheme is proposed for multi-camera dataset. To overcome the high-dimensionality issue, incurred due to multi-camera data, the low-dimensional representation based on Motion History Image (MHI) was extracted. A single MHI is computed for each view/action video. For efficient description of MHIs Histograms of Oriented Gradients (HOG) are employed. Finally the classification of HOG based description of MHIs is based on Nearest Neighbor (NN) classifier. The proposed method does not employ feature fusion for multi-view data and therefore this method does not require a fixed number of cameras setup during training and testing stages. The proposed method is suitable for multi-view as well as single view dataset as no feature fusion is used. Experimentation results on multi-view MuHAVi-14 and MuHAVi-8 datasets give high accuracy rates of 92.65% and 99.26% respectively using Leave-One-Sequence-Out (LOSO) cross validation technique as compared to similar state-of-the-art approaches. The proposed method is computationally efficient and hence suitable for real-time action recognition systems., S.A. Velastin acknowledges funding from the Universidad Carlos III de Madrid, the European Union's Seventh Framework Programme for research, technological development and demonstration under grant agreement n° 600371, el Ministerio de Economia y Competitividad (COFUND2013-51509) and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Efficient Charge Separation of Cold Charge-Transfer States in Organic Solar Cells Through Incoherent Hopping

  • Athanasopoulos, Stavros
  • Tscheuschner, Steffen
  • Bässler, Heinz
  • Köhler, Anna
We demonstrate that efficient and nearly field-independent charge separation of electron hole pairs in organic planar heterojunction solar cells can be described by an incoherent hopping mechanism. Using kinetic Monte Carlo simulations that include the effect of on-chain delocalization as well as entropic contributions, we simulate the dissociation of the charge-transfer state in polymer fullerene bilayer solar cells. The model further explains experimental results of almost field independent charge separation in bilayers of molecular systems with fullerenes and provides important guidelines at the molecular level for maximizing the efficiencies of organic solar cells. Thus, utilizing coherent phenomena is not necessarily required for highly efficient charge separation in organic solar cells., This project has received funding from the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development, and demonstration under Grant Agreement No. 600371, el Ministerio de Economı́a, Industria y Competitividad (COFUND2014-51509), el Ministerio de Educación, cultura y Deporte (CEI-15-17), and Banco Santander. We also acknowledge additional funding from the German Research Foundation DFG (GRK1640) and the Bavarian University Centre for Latin America (BAYLAT).
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

PEO: An immobile solvent?

  • Thiam, A.
  • Martínez Cisneros, Cynthia Susana
  • Molmeret, Y.
  • Iojoiu, C.
  • Sánchez, Jean Yves
Despite used for half a century as host for salt-polymer complexes, PEO is still not a fossil and due to its availability, remains regularly used as a reference in solvent-free polymer electrolytes and related electrochemical cells. Often qualified as macromolecular solvent or immobile solvent, its drawbacks (crystallinity, mechanical strength) are well identified. On the other hand, its electrolyte conductivity maxima are considered as the best possible in absence of molecular solvents or ionic liquids. The comparison of PEO/LiTFSI based on raw PEO and ultrafiltrated one, shows unambiguously the impact of unentangled oligomers not only on ionic transport but also on mechanical behavior. Conductivity, cationic transference numbers and storage modulus data go in the same direction and the cationic conductivity (O/Li = 30) is divided by 2, following PEO purification., Jean-Yves Sanchez acknowledges the CONEX Programme, funding received from Universidad Carlos III de Madrid, the European Union's Seventh Framework Programme for research, technological development and demonstration (Grant agreement nº 600371), Spanish Ministry of Economy and Competitiveness (COFUND2013-40258) and Banco Santander. Amadou Thiam acknowledges ANR for his fellowship. Yannick Molméret acknowledges KICINNO Energy for the granting of his post-doc fellowship, in the frame of the project PENLiB coordinated by Prof. Jean-Yves Sanchez.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Rey's Auditory Verbal Learning Test scores can be predicted from whole brain MRI in Alzheimer's disease

  • Moradi, Elaheh
  • Hallikainen, Ilona
  • Hänninen, Tuomo
  • Tohka, Jussi
Rey's Auditory Verbal Learning Test (RAVLT) is a powerful neuropsychological tool for testing episodic memory, which is widely used for the cognitive assessment in dementia and pre-dementia conditions. Several studies have shown that an impairment in RAVLT scores reflect well the underlying pathology caused by Alzheimer's disease (AD), thus making RAVLT an effective early marker to detect AD in persons with memory complaints. We investigated the association between RAVLT scores (RAVLT Immediate and RAVLT Percent Forgetting) and the structural brain atrophy caused by AD. The aim was to comprehensively study to what extent the RAVLT scores are predictable based on structural magnetic resonance imaging (MRI) data using machine learning approaches as well as to find the most important brain regions for the estimation of RAVLT scores. For this, we built a predictive model to estimate RAVLT scores from gray matter density via elastic net penalized linear regression model. The proposed approach provided highly significant cross-validated correlation between the estimated and observed RAVLT Immediate (R = 0.50) and RAVLT Percent Forgetting (R = 0.43) in a dataset consisting of 806 AD, mild cognitive impairment (MCI) or healthy subjects. In addition, the selected machine learning method provided more accurate estimates of RAVLT scores than the relevance vector regression used earlier for the estimation of RAVLT based on MRI data. The top predictors were medial temporal lobe structures and amygdala for the estimation of RAVLT Immediate and angular gyrus, hippocampus and amygdala for the estimation of RAVLT Percent Forgetting. Further, the conversion of MCI subjects to AD in 3-years could be predicted based on either observed or estimated RAVLT scores with an accuracy comparable to MRI-based biomarkers., This project has received funding from the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development and demonstration under grant agreement no. 600371, el Ministerio de Economía y Competitividad (COFUND2013-40258), el Ministerio de Educación, Cultura y Deporte (CEI-15-17) and Banco Santander., Publicado
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Sodium polymer electrolytes composed of sulfonated polysulfone and macromolecular/molecular solvents for Na-batteries

  • Martínez Cisneros, Cynthia Susana
  • Antonelli, Claire
  • Levenfeld Laredo, Belén
  • Varez, Alejandro
  • Sánchez, Jean Yves
Polysulfone acidic ionomers have been extensively used as Fuel Cell membranes, mainly because of their mechanical, thermal, chemical and electrochemical stability as well as their excellent film-forming capability. This contribution deals with the development of blends based on polysulfone-sodium sulfonate and macromolecular/molecular solvents, consisting of poly(oxyethylene), POE, and propylene carbonate, PC, respectively. The objectives were to take advantage of both the thermomechanical performances and the macromolecular polyanions provided by the polysulfone ionomer. Combining POE/PC solvents, didn't allow obtaining sufficient blend's conductivities. Nevertheless, the addition of very low amounts of sodium perchlorate led to appreciable conductivities., This work has been supported by Projects funded by the regional government (Comunidad de Madrid through MATERYENER3CM S2013/MIT-2753), the Spanish Government, MICINN (MAT2016-78362-C4-3-R) and Fundación Iberdrola España for the Energy and Environment Research Grants 2016. J-Y Sanchez acknowledges the CONEX Programme, funding received from Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development and demonstration (Grant agreement nº 600371), Spanish Ministry of Economy and Competitiveness (COFUND2013-40258) and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

An exergy-based study on the relationship between costs and environmental impacts in power plants

  • Lara, Yolanda
  • Petrakopoulou, Foteini Konstantina
  • Morosuk, Tatiana
  • Boyano, Alicia
  • Tsatsaronis, George
Exergy-based (exergetic, exergoeconomic and exergoenvironmental) analyses, are used for designing, assessing and improving energy conversion systems. In an exergoeconomic analysis, thermodynamic inefficiencies represented by exergy destruction are used in combination with investment costs to calculate the "cost-optimal" layout of a plant. Analogously, in an exergoenvironmental analysis, the aim is to minimize the total environmental impact of a plant. Until today exergoeconomic and exergoenvironmental analyses have been used as separate and distinct tools and the improvement of a plant has been considered in terms of the reduction of either costs or environmental impact. To simultaneously decrease the investment costs and the component-related (manufacturing or construction -related) environmental impacts, their relationship with exergy destruction must be studied in parallel. This paper examines the relationship between exergoeconomic and exergoenvironmental data under various plant operating conditions. A combined-cycle power plant is analyzed and options for a simultaneous improvement from the thermodynamic, economic and environmental viewpoints are discussed. (C) 2017 Elsevier Ltd. All rights reserved., Tatiana Morosuk gratefully acknowledges the financial support from the Berliner Programm zur Förderung der Chancengleichheit von Frauen in Forschung und Lehre. Fontina Petrakopoulou would like to thank the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development and demonstration (grant agreement nº 600371), the Ministerio de Economía y Competitividad (COFUND2014-51509) and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Contribución de congreso (ConferenceOutput)

Motorcycle detection and classification in urban Scenarios using a model based on Faster R-CNN

  • Espinosa, Jorge E.
  • Velastin Carroza, Sergio Alejandro
  • Branch, John W.
This paper has been presented at: 9th International Conference on Pattern Recognition Systems (ICPRS-18), This paper introduces a Deep Learning Convolutional Neutral Network model based on Faster-RCNN for motorcycle detection and classification on urban environments. The model is evaluated in occluded scenarios where more than 60% of the vehicles present a degree of occlusion. For training and evaluation, we introduce a new dataset of 7500 annotated images, captured under real traffic scenes, using a drone mounted camera. Several tests were carried out to design the network, achieving promising results of 75% in average precision (AP), even with the high number of occluded motorbikes, the low angle of capture and the moving camera. The model is also evaluated on low occlusions datasets, reaching results of up to 92% in AP., S.A. Velastin is grateful to funding received from the Universidad Carlos III de Madrid, the European Union's Seventh Framework Programme for research, technological development and demonstration under grant agreement no. 600371, el Ministerio de Economía y Competitividad (COFUND2013-51509) and Banco Santander. The authors gratefully acknowledge the support of NVIDIA Corporation with the donation of the GPUs used for this research. The data and code used for this work is available upon request from the authors.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Increasing Photovoltaic Performance of an Organic Cationic Chromophore by Anion Exchange

  • Gesevicius, Donatas
  • Neels, Antonia
  • Jenatsch, Sandra
  • Hack, Erwin
  • Viani, Lucas
  • Athanasopoulos, Stavros
  • Nüesch, Frank
  • Heier, Jakob
A symmetrical cyanine dye chromophore is modified with different counteranions to study the effect on crystal packing, polarizability, thermal stability, optical properties, light absorbing layer morphology, and organic photovoltaic (OPV) device parameters. Four sulfonate-based anions and the bulky bistriflylimide anion are introduced to the 2-[5-(1,3-dihydro-1,3,3-trimethyl-2H-indol-2-ylidene)-1,3-pentadien-1-yl]-1,3,3-trimethyl-3H-indolium chromophore using an Amberlyst A26 (OH− form) anion exchanger. Anionic charge distribution clearly correlates with device performance, whereby an average efficiency of 2% was reached in a standard bilayer organic solar. Evidence is given that the negative charge of the anion distributed over a large number of atoms is significantly more important than the size of the organic moieties of the sulfonate charge carrying group. This provides a clear strategy for future design of more efficient cyanine dyes for OPV applications., The authors thank Dr. Matthias Nagel (Empa) for suggestions and Dr. Mohammed Makha (Empa) for his help with the thermal evaporation techniques. The use of the Scanning Probe Microscopy User Laboratory at Empa is gratefully acknowledged. The authors thank the Swiss National Science Foundation for financial support under grant number 160116. This project has also received funding from the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Program for research, technological development and demonstration under grant agreement No. 600371, el Ministerio de Economía, Industria y Competitividad (COFUND2014-51509), el Ministerio de Educación, cultura y Deporte (CEI-15-17), and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

A Unified Deep Framework for Joint 3D Pose Estimation and Action Recognition from a Single RGB Camera

  • Pham, Huy-Hieu
  • Salmane, Houssam
  • Khoudour, Louahdi
  • Crouzil, Alain
  • Velastin Carroza, Sergio Alejandro
  • Zegers, Pablo
We present a deep learning-based multitask framework for joint 3D human pose estimation and action recognition from RGB sensors using simple cameras. The approach proceeds along two stages. In the first, a real-time 2D pose detector is run to determine the precise pixel location of important keypoints of the human body. A two-stream deep neural network is then designed and trained to map detected 2D keypoints into 3D poses. In the second stage, the Efficient Neural Architecture Search (ENAS) algorithm is deployed to find an optimal network architecture that is used for modeling the spatio-temporal evolution of the estimated 3D poses via an image-based intermediate representation and performing action recognition. Experiments on Human3.6M, MSR Action3D and SBU Kinect Interaction datasets verify the effectiveness of the proposed method on the targeted tasks. Moreover, we show that the method requires a low computational budget for training and inference. In particular, the experimental results show that by using a monocular RGB sensor, we can develop a 3D pose estimation and human action recognition approach that reaches the performance of RGB-depth sensors. This opens up many opportunities for leveraging RGB cameras (which are much cheaper than depth cameras and extensively deployed in private and public places) to build intelligent recognition systems., Sergio A. Velastin is grateful for funding received from the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development and demonstrationunder grant agreement N 600371, el Ministerio de Economía, Industria y Competitividad (COFUND2013-51509) el Ministerio de Educación, Cultura y Deporte (CEI-15-17) and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Cost-based design optimization of the heat exchangers in a parabolic trough power plant

  • González Gómez, Pedro Ángel
  • Petrakopoulou, Foteini Konstantina
  • Villa Briongos, Javier
  • Santana Santana, Domingo José
This paper addresses two important concerns of the design of steam generators of parabolic trough power plants: cost minimization and component reliability. A thorough economic analysis of the heat exchangers of the steam generator and oil-to-salt heat exchangers of a 50 MWe parabolic trough power plant is presented. The heat exchanger design is realized following TEMA standards and optimized using a genetic algorithm. Two design strategies are compared: the minimization of the total heat transfer area and the minimization of the total annualized cost. It is seen that the second approach provides substantial savings over the lifetime of the plant. The economic analysis reveals a global optimum with an outlet temperature of the heat transfer fluid of 293 degrees C and an evaporator pinch point of 4.85 degrees C. The best design of the steam generator consists of a TEMA-H shell superheater and preheater and a TEMA-F shell reheater. The best design of the oil-to-salt heat exchangers includes six TEMA-F shell heat exchangers in series, with a log mean temperature difference of 7 degrees C and the molten salt on the shell-side. Lastly, a TEMA-X recirculation evaporator is proposed with a considerably reduced wall thickness when compared to a kettle evaporator. (C) 2017 Elsevier Ltd. All rights reserved., This work was supported by project ENE-2015-69486-R of the Ministerio de Economía y Competitividad. F.P. would like to thank the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development and demonstration (grant agreement no 600371), the Ministerio de Economía y Competitividad (COFUND2014-51509) the Ministerio de Educación, Cultura y Deporte (CEI-15-17) and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Contribución de congreso (ConferenceOutput)

People Detection and Pose Classification Inside a Moving Train Using Computer Vision

  • Velastin Carroza, Sergio Alejandro
  • Gómez-Lira, Diego A.
This paper has been presented at : 5th International Visual Informatics Conference (IVIC 2017), Also part of the Image Processing, Computer Vision, Pattern Recognition, and Graphics book sub series (LNIP, volume 10645), The use of surveillance video cameras in public transport is increasingly regarded as a solution to control vandalism and emergency situations. The widespread use of cameras brings in the problem of managing high volumes of data, resulting in pressure on people and resources. We illustrate a possible step to automate the monitoring task in the context of a moving train (where popular background removal algorithms will struggle with rapidly changing illumination). We looked at the detection of people in three possible postures: Sat down (on a train seat), Standing and Sitting (half way between sat down and standing). We then use the popular Histogram of Oriented Gradients (HOG) descriptor to train Support Vector Machines to detect people in any of the predefined postures. As a case study, we use the public BOSS dataset. We show different ways of training and combining the classifiers obtaining a sensitivity performance improvement of about 12% when using a combination of three SVM classifiers instead of a global (all classes) classifier, at the expense of an increase of 6% in false positive rate. We believe this is the first set of public results on people detection using the BOSS dataset so that future researchers can use our results as a baseline to improve upon., The work described here was carried out as part of the OBSERVE project funded by the Fondecyt Regular Program of Conicyt (Chilean Research Council for Science and Technology) under grant no. 1140209. S.A. Velastin is grateful to funding received from the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development and demonstration under grant agreement no. 600371, el Ministerio de Economía y Competitividad (COFUND2013-51509) and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Multiple Batches of Motion History Images (MB-MHIs) for Multi-view Human Action Recognition

  • Naeem, Hajra
  • Murtaza, Fiza
  • Yousaf, Muhammad Haroon
  • Velastin Carroza, Sergio Alejandro
The recognition of human actions recorded in a multi-camera environment faces the challenging issue of viewpoint variation. Multi-view methods employ videos from different views to generate a compact view-invariant representation of human actions. This paper proposes a novel multi-view human action recognition approach that uses multiple low-dimensional temporal templates and a reconstruction-based encoding scheme. The proposed approach is based upon the extraction of multiple 2D motion history images (MHIs) of human action videos over non-overlapping temporal windows, constructing multiple batches of motion history images (MB-MHIs). Then, two kinds of descriptions are computed for these MHIs batches based on (1) a deep residual network (ResNet) and (2) histogram of oriented gradients (HOG) to effectively quantify a change in gradient. ResNet descriptions are average pooled at each batch. HOG descriptions are processed independently at each batch to learn a class-based dictionary using a K-spectral value decomposition algorithm. Later, the sparse codes of feature descriptions are obtained using an orthogonal matching pursuit approach. These sparse codes are average pooled to extract encoded feature vectors. Then, encoded feature vectors at each batch are fused to form a final view-invariant feature representation. Finally, a linear support vector machine classifier is trained for action recognition. Experimental results are given on three versions of a multi-view dataset: MuHAVi-8, MuHAVi-14, and MuHAVi-uncut. The proposed approach shows promising results when tested for a novel camera. Results on deep features indicate that action representation by MB-MHIs is more view-invariant than single MHIs., Muhammad Haroon Yousaf has received funding from Higher Education Commission, Pakistan, for Swarm Robotics Lab under National Centre for Robotics and Automation (NCRA). S.A. Velastin is grateful to funding received from the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development, and demonstration under Grant Agreement No. 600371, el Ministerio de Economía y Competitividad (COFUND2013-51509), and Banco Santander. The authors also acknowledge support from the Directorate of ASR&TD, University of Engineering and Technology, Taxila, Pakistan, and of Nvidia Corporation for its donation of GPU equipment.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

EspiNet V2: a region based deep learning model for detecting motorcycles in urban scenarios, EspiNet V2: un modelo basado en regiones de aprendizaje profundo para detectar motocicletas en escenarios urbanos

  • Espinosa, Jorge E.
  • Velastin Carroza, Sergio Alejandro
  • Branch Bedoya, John William
This paper presents "EspiNet V2" a Deep Learning model, based on the region-based detector Faster R-CNN. The model is used for the detection of motorcycles in urban environments, where occlusion is likely. For training, two datasets are used: the Urban Motorbike Dataset (UMD-10K) of 10,000 annotated images, and the new SMMD (Secretaría de Movilidad Motorbike Dataset), of 5,000 images captured from the Traffic Control CCTV System in Medellín (Colombia). Results achieved on the UMD-10K dataset reach 88.8% in average precision (AP) even when 60% motorcycles were occluded, and the images were captured from a low angle and a moving camera. Meanwhile, an AP of 79.5% is reached for SSMD. EspiNet V2 outperforms popular models such as YOLO V3 and Faster R-CNN (VGG16 based) trained end-to-end for those datasets., Este artículo presenta "EspiNet V2", un modelo de aprendizaje profundo, fundamentado en el detector basado regiones Faster R-CNN. El modelo es usado para la detección de motocicletas en entornos urbanos, donde se presenta algún nivel de oclusión. Para el entrenamiento de dicho modelo, se utilizaron dos conjuntos de datos: el conjunto de datos de motocicletas urbanas (UMD-10K) que cuenta con 10,000 imágenes anotadas, y el nuevo conjunto de datos de motos de la Secretaría de Movilidad (SMMD), con 5,000 imágenes capturadas obtenidas del Sistema CCTV de Control de Tráfico de la ciudad de Medellín (Colombia). Los resultados obtenidos en el conjunto de datos UMD-10K alcanzan el 88.8% en precisión promedio (AP), incluso con niveles de oclusión de un 60 %, utilizando imágenes capturadas desde un ángulo bajo y desde una cámara en movimiento. Por otro lado se alcanza un AP de 79.5 % para conjunto de datos de motos de la Secretaría de Movilidad (SMMD). EspiNet V2 supera modelos populares como YOLO V3 y Faster R-CNN (basado en VGG16), siendo estos entrenados de extremo a extremo utilizando los conjuntos de datos mencionados., Sergio A. Velastin is grateful for funding received from the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for Research, Technological Development and demonstration under grant agreement N. 600371, el Ministerio de Economia, Industria y Competitividad (COFUND2013-51509) el Ministerio de Educación, Cultura y Deporte (CEI-15-17) and Banco Santander. This work was partially supported by COLCIENCIAS project: Reduccion de Emisiones Vehiculares Mediante el Modelado y Gestion Optima de Trafico en Areas Metropolitanas - Caso Medellin - Area Metropolitana del Valle de Aburra, codigo 111874558167, CT 049-2017. Universidad Nacional de Colombia- Politécnico Colombiano Jaime Isaza Cadavid. Proyecto HERMES 25374. The authors also gratefully acknowledge the support of NVIDIA Corporation with the donation of the two GPUs used for this research.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

PMHI: Proposals From Motion History Images for Temporal Segmentation of Long Uncut Videos

  • Murtaza, Fiza
  • Yousaf, Muhammad Haroon
  • Velastin Carroza, Sergio Alejandro
This letter proposes a method for the generation of temporal action proposals for the segmentation of long uncut video sequences. The presence of consecutive multiple actions in video sequences makes the temporal segmentation a challenging problem due to the unconstrained nature of actions in space and time. To address this issue, we exploit the nonaction segments present between the actual human actions in uncut videos. From the long uncut video, we compute the energy of consecutive nonoverlapping motion history images (MHIs), which provides spatiotemporal information of motion. Our proposals from MHIs (PMHI) are based on clustering the MHIs into actions and nonaction segments by detecting minima from the energy of MHIs. PMHI efficiently segments the long uncut videos into a small number of nonoverlapping temporal action proposals. The strength of PMHI is that it is unsupervised, which alleviates the requirement for any training data. Our temporal action proposal method outperforms the existing proposal methods on the Multi-view Human Action video (MuHAVi)-uncut and Computer Vision and Pattern recognition (CVPR) 2012 Change Detection datasets with an average recall rate of 86.1% and 86.0%, respectively., Sergio A Velastin acknowledges funding by the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development and demonstration under grant agreement nº 600371, el Ministerio de Economía y Competitividad (COFUND2013-51509) and Banco Santander
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Contribución de congreso (ConferenceOutput)

Learning and Recognizing Human Action from Skeleton Movement with Deep Residual Neural Networks

  • Pham, Huy-Hieu
  • Khoudour, Louahdi
  • Crouzil, Alain
  • Zegers, Pablo
  • Velastin Carroza, Sergio Alejandro
This paper has been presented at 8th International Conference of Pattern Recognition Systems (ICPRS 2017)., Automatic human action recognition is indispensable for almost artificial intelligent systems such as video surveillance, human-computer interfaces, video retrieval, etc. Despite a lot of progresses, recognizing actions in a unknown video is still a challenging task in computer vision. Recently, deep learning algorithms has proved its great potential in many vision-related recognition tasks. In this paper, we propose the use of Deep Residual Neural Networks (ResNets) to learn and recognize human action from skeleton data provided by Kinect sensor. Firstly, the body joint coordinates are transformed into 3D-arrays and saved in RGB images space. Five different deep learning models based on ResNet have been designed to extract image features and classify them into classes. Experiments are conducted on two public video datasets for human action recognition containing various challenges. The results show that our method achieves the state-of-the-art performance comparing with existing approaches, This work was supported by the Cerema Research Center and Universidad Carlos III de Madrid. Sergio A. Velastin has received funding from the European Unions Seventh Framework Programme for Research, Technological Development and demonstration under grant agreement No 600371, el Ministerio de Economía, Industria y Competitividad (COFUND2013-51509) el Ministerio de Educación, cultura y Deporte (CEI-15-17) and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Theoretical study of equilibrium and nonequilibrium exciton dynamics in disordered semiconductors

  • Ansari-Rad, Mehdi
  • Athanasopoulos, Stavros
We develop a temperature-dependent theory for singlet exciton hopping transport in disordered semiconductors. It draws on the transport level concept within a Förster transfer model and bridges the gap in describing the transition from equilibrium to nonequilibrium time-dependent spectral diffusion.We test the validity range of the developed model using kinetic Monte Carlo simulations and find agreement over a broad range of temperatures. It reproduces the scaling of the diffusion length and spectral shift with the dimensionless disorder parameter and describes in a unified manner the transition from equilibrium to nonequilibrium transport regime. We find that the diffusion length in the nonequilibrium regime does not scale with the the third power of the Förster radius. The developed theory provides a powerful tool for interpreting time-resolved and steady state spectroscopy experiments in a variety of disordered materials, including organic semiconductors and colloidal quantum dots., This project has received funding from the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development, and demonstration under Grant Agreement No. 600371, el Ministerio de Economía, Industria y Competitividad (COFUND2014-51509), el Ministerio de Educación, cultura y Deporte (CEI-15-17), and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Vectors of Temporally Correlated Snippets for Temporal Action Detection

  • Murtaza, Fiza
  • Haroon Yousaf, Muhammad
  • Velastin Carroza, Sergio Alejandro
  • Qian, Yu
Detection of human actions in long untrimmed videos is an important but challenging task due to the unconstrained nature of actions present in untrimmed videos. We argue that untrimmed videos contain multiple snippets from actions and the background classes having significant correlation with each other, which results in imprecise detection of start-end times for action regions. In this work, we propose Vectors of Temporally Correlated Snippets (VTCS) which addresses this problem by finding the snippet-centroids from each class which are discriminant for their own class. For each untrimmed video, non-overlapping snippets are temporally correlated with the snippet-centroids using VTCS encoding to find the action proposals. We evaluate the performance of VTCS on the Thumos14 and ActivityNet datasets. For Thumos14, VTCS achieves a significant gain in mean Average Precision (mAP) at temporal Intersection over Union (tIoU) threshold 0.5, improving from 41.5% to 44.3%. For the sports-subset of ActivityNet dataset, VTCS obtains 38.5% mAP @0.5 tIoU threshold., We gratefully acknowledge the funding for Swarm Robotics Lab under National Centre for Robotics and Automation, Pakistan. We also acknowledge the support of NVIDIA Corporation with the donation of the TITAN Xp GPU used for this research. Sergio A Velastin acknowledges funding by the Universidad Carlos III de Madrid, the European Unions Seventh Framework Programme for research, technological development and demonstration under grant agreement no. 600371, el Ministerio de Economía y Competitividad (COFUND2013-51509) and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Non-woven polyaramid porous membranes as separators for Li-ion batteries?

  • Martínez Cisneros, Cynthia Susana
  • Antonelli, Claire
  • Levenfeld Laredo, Belén
  • Lumbier Álvarez, Alejandro
  • Perez-Flores, Juan Carlos
  • Santos-Mendez, Antonio
  • Kuhn, Alois
  • Sánchez, Jean Yves
Macroporous separators play a crucial role regarding safety in current Li-ion batteries. Most separators used in battery applications are based on polyolefin and present shrinkage and a decrease in mechanical properties when used at high temperatures, both detrimental in the battery performance. In search of more suitable alternatives that render to more stable and safer batteries, in this work, non-woven separators based on polyacrylonitrile blended with cellulose and para-aramid fibers are systematically investigated. This study has been carried out in terms of microstructure, mechanical properties, ionic conductivity and thermal and electrochemical stability (using Li4Ti5O12/LiCoO2 full cells with a nominal potential of ≈2.5 V). Although conductivity values of separators are somewhat modest, the electrochemical performance developed when used in Li4Ti5O12/LiCoO2 cells are, at moderate C rates, comparable to commercial Celgard©2400 separator. At high C rates, in particular at 2C, DwG40 exhibits much higher capacities than the whole of the separators, including Celgard®2400. This study responds to the continuous need reflected not only by the scientific community but also by the industrial one when new materials for electrolytes and electrodes should be tested, since there is a lack of reports characterizing such elements in the literature., We thank "Agencia Estatal de Investigación/Fondo Europeo de Desarrollo Regional" (FEDER/UE) for funding the projects MAT2016-78362-C4-3-R, MAT2016-78362-C4-1-R, PID2019-106662RB-C41 and PID2019-106662RB-C44. The regional government of Comunidad de Madrid is acknowledged for funding the project MATERYENER3CM S2013/MIT-2753 and the predoctoral contract PEJD-2019-PRE/IND-16924. J-Y Sanchez acknowledges the CONEX Programme, funding received from Universidad Carlos III de Madrid, the European Union's Seventh Framework Programme for research, technological development and demonstration (Grant agreement n° 600371), Spanish Ministry of Economy and Competitiveness (COFUND2013-40258) and Banco Santander. The authors would like to acknowledge DreamWeaver for samples supply.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Contribución de congreso (ConferenceOutput)

Inter and Intra Class Correlation Analysis (IIcCA) for Human Action Recognition in Realistic Scenarios

  • Nazir, Saima
  • Yousaf, Muhammad Haroon
  • Velastin Carroza, Sergio Alejandro
This paper has been presented at : 8th International Conference of Pattern Recognition Systems (ICPRS 2017), Human action recognition in realistic scenarios is an important yet challenging task. In this paper we propose a new method, Inter and Intra class correlation analysis (IICCA), to handle inter and intra class variations observed in realistic scenarios. Our contribution includes learning a class specific visual representation that efficiently represents a particular action class and has a high discriminative power with respect to other action classes. We use statistical measures to extract visual words that are highly intra correlated and less inter correlated. We evaluated and compared our approach with state-of-the-art work using a realistic benchmark human action recognition dataset., S.A. Velastin has received funding from the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development and demonstration under grant agreement no 600371, the Ministerio de Economía, Industria y Competitividad (COFUND2013-51509) the Ministerio de Educación, cultura y Deporte (CEI-15-17) and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Contribución de congreso (ConferenceOutput)

Human Action Recognition using Multi-Kernel Learning for Temporal Residual Network

  • Nazir, Saima
  • Qian, Yu
  • Yousaf, Muhammad Haroon
  • Velastin Carroza, Sergio Alejandro
  • Izquierdo, Ebroul
  • Vazquez, Eduard
This paper has been presented at the 14th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications., Deep learning has led to a series of breakthrough in the human action recognition field. Given the powerful representational ability of residual networks (ResNet), performance in many computer vision tasks including human action recognition has improved. Motivated by the success of ResNet, we use the residual network and its variations to obtain feature representation. Bearing in mind the importance of appearance and motion information for action representation, our network utilizes both for feature extraction. Appearance and motion features are further fused for action classification using a multi-kernel support vector machine (SVM).We also investigate the fusion of dense trajectories with the proposed network to boost up the network performance. We evaluate our proposed methods on a benchmark dataset (HMDB-51) and results shows the multi-kernel learning shows the better performance than the fusion of classification score from deep network SoftMax layer. Our proposed method also shows good performance as compared to the recent state-of-the-art methods., Sergio A. Velastin has received funding from the Universidad Carlos III de Madrid, the European Unions Seventh Framework Programme for research, technological development and demonstration under grant agreement n◦ 600371, el Ministerio de Economía, Industria y Competitividad (COFUND2013-51509) el Ministerio de Educación, cultura y Deporte (CEI-15-17) and Banco Santander. Authors also acknowledge support from the Higher Education Commission, Pakistan.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Exergetic analysis and dynamic simulation of a solar-wind power plant with electricity storage and hydrogen generation

  • Petrakopoulou, Foteini Konstantina
  • Alexander, Robinson
  • Loizidou, María
The ambitious vision of off-grid renewable energy autonomy of remote regions has yet to come to fruition. The development of comprehensive energy production systems would be needed to achieve such a goal. This study consists of the simulation and exergetic evaluation of a novel hybrid power plant for stand-alone operation aiming to provide electricity autonomy of a Mediterranean island. The considered power plant is simulated dynamically over an annual cycle and accounts for both energy input fluctuations and electricity surplus. The plant combines a photovoltaic array with wind turbines for energy input, coupled with electricity storage and a hydrogen generation facility to stabilize the power output of the plant. Unlike other similar studies, the energy system presented here relies on real-case weather and demand data of a relatively large remote community and is optimized to ensure continuous operation even under extreme conditions. It is seen that this stand-alone hybrid power plant constitutes a robust and secure alternative to the current conventional energy situation; the combined renewable technologies succeed in complementing each other and offer stable performance throughout the year without the requirement of additional support by fossil fuels. The mean annual exergetic efficiency of the plant is found to be 17.9%, producing approximately 25,000 MWh of electricity per year, along with a secondary product (hydrogen) produced in the electrolyzer of the plant. Although this additional product is associated with additional investment cost, it offers the possibility to stabilize the power plant's performance and can be used as an additional source of financial income for the community. (C) 2015 Elsevier Ltd. All rights reserved., The authors would like to acknowledge the Marie Curie Actions PEOPLE-2012 IEFGENERGIS- 332028 and COFUND CONEX funded by FP7.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Corporate agency in European cinema: A special issue of Studies in European cinema

  • Meir, Christopher Tunstil
Introduction to the special issue 16(1) of Studies in European Cinema, This introduction to the special issue grows out of a larger research project which has received funding from the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development and demonstration under grant agreement nº 600371, el Ministerio de Economía, Industria y Competitividad (COFUND2014-51509) el Ministerio de Educación, cultura y Deporte (CEI-15-17) and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Contribución de congreso (ConferenceOutput)

Bag of Deep Features for Instructor Activity Recognition in Lecture Room

  • Nida, Nudrat
  • Yousaf, Muhammad Haroon
  • Irtaza, Aun
  • Velastin Carroza, Sergio Alejandro
This paper has been presented at : 25th International Conference on MultiMedia Modeling (MMM2019), This research aims to explore contextual visual information in the lecture room, to assist an instructor to articulate the effectiveness of the delivered lecture. The objective is to enable a self-evaluation mechanism for the instructor to improve lecture productivity by understanding their activities. Teacher’s effectiveness has a remarkable impact on uplifting students performance to make them succeed academically and professionally. Therefore, the process of lecture evaluation can significantly contribute to improve academic quality and governance. In this paper, we propose a vision-based framework to recognize the activities of the instructor for self-evaluation of the delivered lectures. The proposed approach uses motion templates of instructor activities and describes them through a Bag-of-Deep features (BoDF) representation. Deep spatio-temporal features extracted from motion templates are utilized to compile a visual vocabulary. The visual vocabulary for instructor activity recognition is quantized to optimize the learning model. A Support Vector Machine classifier is used to generate the model and predict the instructor activities. We evaluated the proposed scheme on a self-captured lecture room dataset, IAVID-1. Eight instructor activities: pointing towards the student, pointing towards board or screen, idle, interacting, sitting, walking, using a mobile phone and using a laptop, are recognized with an 85.41% accuracy. As a result, the proposed framework enables instructor activity recognition without human intervention., Sergio A Velastin has received funding from the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development and demonstration under grant agreement no 600371, el Ministerio de Economía, Industria y Competitividad (COFUND2014-51509) el Ministerio de Educación, Cultura y Deporte (CEI-15-17) and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Evaluating a bag-of-visual features approach using spatio-temporal features for action recognition

  • Velastin Carroza, Sergio Alejandro
  • Nazir, Saima
  • Yousaf, Muhammad Haroon
The detection of the spatial-temporal interest points has a key role in human action recognition algorithms. This research work aims to exploit the existing strength of bag-of-visual features and presents a method for automatic action recognition in realistic and complex scenarios. This paper provides a better feature representation by combining the benefit of both a well-known feature detector and descriptor i.e. the 3D Harris space-time interest point detector and the 3D Scale-Invariant Feature Transform descriptor. Finally, action videos are represented using a histogram of visual features by following the traditional bag-of-visual feature approach. Apart from video representation, a support vector machine (SVM) classifier is used for training and testing. A large number of experiments show the effectiveness of our method on existing benchmark datasets and shows state-of-the-art performance. This article reports 68.1% mean Average Precision (mAP), 94% and 91.8% average accuracy for Hollywood-2, UCF Sports and KTH datasets respectively. (C) 2018 Elsevier Ltd. All rights reserved., Sergio A Velastin has received funding from the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development and demonstration under grant agreement no. 600371, el Ministerio de Economía, Industria y Competitividad (COFUND2013-51509) el Ministerio de Educación, cultura y Deporte (CEI-15-17) and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Capítulo o parte de un libro (BookPart)

Against human exceptionalism: environmental ethics and the machine question

  • Laukyte, Migle
This paper offers an approach for addressing the question of how to deal with artificially intelligent entities, such as robots, mindclones, androids, or any other entity having human features. I argue that to this end we can draw on the insights offered by environmental ethics, suggesting that artificially intelligent entities ought to be considered not as entities that are extraneous to the human social environment, but as forming an integral part of that environment. In making this argument I take a radical strand of environmental ethics, namely, Deep Ecology, which sees all entities as existing in an inter-relational environment: I thus reject any “firm ontological divide in the field of existence” (Fox W, Deep ecology: A new philosophy of our time? In: Light A, Rolston III H (eds) Environmental ethics: An anthologyBlackwell, Oxford, 252–261, 2003) and on that basis I introduce principles of biospherical egalitarianism, diversity, and symbiosis (Naess A, Inquiry 16(1):95–100, 1973). Environmental ethics makes the case that humans ought to “include within the realms of recognition and respect the previously marginalized and oppressed” ((Gottlieb RS, Introduction. In: Merchant C (ed) Ecology. Humanity Books, Amherst, pp ix–xi, 1999)). I thus consider (a) whether artificially intelligent entities can be described along these lines, as somehow “marginalized” or “oppressed,” (b) whether there are grounds for extending to them the kind of recognition that such a description would seem to call for, and (c) whether Deep Ecology could reasonably be interpreted in such a way that it apply to artificially intelligent entities., This paper is part of the project ALLIES (Artificially Intelligent Entities: Their Legal Status in the Future) that has received funding from the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development and demonstration under grant agreement n◦ 600371, el Ministerio de Economía, Industria y Competitividad (COFUND2014-51509) el Ministerio de Educación, cultura y Deporte (CEI-15-17) and Banco Santander.”
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Simulation and analysis of a stand-alone solar-wind and pumped-storage hydropower plant

  • Petrakopoulou, Foteini Konstantina
  • Robinson, Alexander
  • Loizidou, María
This work presents the simulation and evaluation of a renewable hybrid power plant for off-grid fully autonomous operation on an intermediate-sized island in the Aegean Sea. A stand-alone energy system including storage facilities is simulated, optimized and analyzed relying on real-case weather and demand data of a relatively large remote community. Optimization of the power plant structure shows that to ensure continuous off-grid energy generation, even under extreme conditions, the combination of more than one renewable technology is required. The hybrid power plant consists of a pumped-storage hydropower plant, photovoltaic cells and wind turbines. Energy surplus of the power plant is used in the incorporated electrolyzer to generate a secondary product, hydrogen. Robust operation of the plant results in 48% of the energy generated stemming from the photovoltaic system and 52% from the wind turbines. The pumped-storage hydropower plant has a mean annual power output of 1.0 MW. The total mean annual efficiency of the hybrid plant is found to be 14.4%. Although stand-alone operation was achieved with the proposed plant, this requirement led to net energy output restrictions, capacity oversizing and large storage facilities. (C) 2015 Elsevier Ltd. All rights reserved., The authors would like to acknowledge the Marie Curie Actions PEOPLE-2012-IEFGENERGIS-332028 and COFUND CONEX, funded by FP7.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Ultrafast Non-Forster Intramolecular Donor Acceptor Excitation Energy Transfer

  • Athanasopoulos, Stavros
  • Alfonso Hernández, Laura
  • Beljonne, David
  • Fernández Alberti, Sebastián
  • Tretiak, Sergei
Ultrafast intramolecular electronic energy transfer in a conjugated donor-acceptor system is simulated using nonadiabatic excited-state molecular dynamics. After initial site-selective photoexcitation of the donor, transition density localization is monitored throughout the S-2 -> S-1 internal conversion process, revealing an efficient unidirectional donor acceptor energy-transfer process. Detailed analysis of the excited state trajectories uncovers several salient features of the energy-transfer dynamics. While a weak temperature dependence is observed during the entire electronic energy relaxation, an ultrafast initially temperature-independent process allows the molecular system to approach the S-2-S-1 potential energy crossing seam within the first ten femtoseconds. Efficient energy transfer occurs in the absence of spectral overlap between the donor and acceptor units and is assisted by a transient delocalization phenomenon of the excited-state wave function acquiring Frenkel-exciton character at the moment of quantum transition., This project has received funding from the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development and demonstration under grant agreement n° 600371, el Ministerio de Economia y Competitividad (COFUND2014-51509), el Ministerio de Educación, cultura y Deporte (CEI-15-17) and Banco Santander. This work was partially supported by CONICET, UNQ, ANPCyT (PICT-2014-2662). We also acknowledge support of the Center for Integrated Nano-technology (CINT), a U.S. Department of Energy, Office of Basic Energy Sciences user facility, as well as additional funding from the Bavarian University Centre for Latin America (BAYLAT). The work in Mons is supported by BELSPO through the PAI P6/27 Functional Supramolecular Systems project and by the Belgian National Fund for Scientific Research FNRS/F.R.S. DB is a Research Director of FNRS.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

The Social Perspective on the Renewable Energy Autonomy of Geographically Isolated Communities: Evidence from a Mediterranean Island

  • Petrakopoulou, Foteini Konstantina
The use of renewable energy sources can provide a path toward sustainable developmentand regional energy independence. In particular, renewable-based energy autonomy constitutes aviable option in remote areas. This work presents a survey on the use of renewable energy as part of anenergy autonomy plan on a Mediterranean island. The study also included personal communicationswith residents and local community leaders. The results show an overall positive attitude towardrenewable energy applications. The majority of the respondents support the implementation ofrenewable-based, small-scale projects corresponding to local energy autonomy scenarios. They are,furthermore, convinced that a wider use of renewable technologies can reduce the environmentalimpact of conventional fuels. However, although people are aware of technologies widely used onthe island, they are much less so when it comes to less prominent technologies (wave energy, fuelcells, etc.). People tend to be more open to installations of solar, wind and geothermal energy, whilegenerally they dislike nuclear and coal power plants. Lastly, the majority of the respondents believethat local policies on energy issues should change, while they also perceive the lack of political willas one of the most important obstacles to the implementation of renewable technologies.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Contribución de congreso (ConferenceOutput)

A Deep Learning Approach for Real-Time 3D Human Action Recognition from Skeletal Data

  • Hieu Pham, Huy
  • Salmane, Houssam
  • Khoudour, Louahdi
  • Crouzil, Alain
  • Zegers, Pablo
  • Velastin Carroza, Sergio Alejandro
This paper has been published at the Proceedings of 16th International Conference on Image Analysis and Recognition, Contains Supplementary material, We present a new deep learning approach for real-time 3D human action recognition from skeletal data and apply it to develop a vision-based intelligent surveillance system. Given a skeleton sequence, we propose to encode skeleton poses and their motions into a single RGB image. An Adaptive Histogram Equalization (AHE) algorithm is then applied on the color images to enhance their local patterns and generate more discriminative features. For learning and classification tasks, we design Deep Neural Networks based on the Densely Connected Convolutional Architecture (DenseNet) to extract features from enhanced-color images and classify them into classes. Experimental results on two challenging datasets show that the proposed method reaches state-of-the-art accuracy, whilst requiring low computational time for training and inference. This paper also introduces CEMEST, a new RGB-D dataset depicting passenger behaviors in public transport. It consists of 203 untrimmed real-world surveillance videos of realistic normal and anomalous events. We achieve promising results on real conditions of this dataset with the support of data augmentation and transfer learning techniques. This enables the construction of real-world applications based on deep learning for enhancing monitoring and security in public transport., This research was supported by the Cerema, France. Sergio A. Velastin is grateful for funding from the Universidad Carlos III de Madrid, the EU’s 7th Framework Programme for Research, Technological Development and demonstration (grant 600371), Ministerio de Economia, Industria y Competitividad (COFUND2013- 51509), Ministerio de Educación, cultura y Deporte (CEI-15-17) and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Development of sodium-conducting polymer electrolytes: comparison between film-casting and films obtained via green processes

  • Martínez Cisneros, Cynthia Susana
  • Levenfeld Laredo, Belén
  • Varez, Alejandro
  • Sánchez, Jean Yves
Sodium-conducting solvent-free polymer electrolytes based on commercially available and inexpensive materials poly(oxyethylene), POE, and three different sodium salts (Nal, NaCF3SO3 and NaClO4) were prepared and exhaustively characterized. In order to minimize the environmental impact related to conventional film processing based on casting, a combination of lyophilization and hot-pressing was successfully applied. Contrary to film-casting, this new approach led to very homogeneous and pore-free films. This study suggests the obtained polymer electrolyte films as a promising route to enhance not only ionic conductivity but also mechanical properties. Furthermore, a preliminary work on salt blends hosted by POE shows that they strongly decrease melting point and crystallinity of the polymer electrolytes and paves the way for enhanced sodium-conducting materials. (C) 2016 Elsevier Ltd. All rights reserved., This work has been supported by Projects funded by the regional government (Comunidad de Madrid through MATERYENER3CM S2013/MIT-2753) and the Spanish Government, MICINN (MAT2013-46452-C4-3R). J-Y Sanchez acknowledges Universidad Carlos III de Madrid for awarding him a European Chair of Excellence and the CONEX Programme funding, received from Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development and demonstration (Grant agreement n 600371), Spanish Ministry of Economy and Competitiveness (COFUND2013-40258) and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Multi-view human action recognition using 2D motion templates based on MHIs and their HOG description

  • Murtaza, Fiza
  • Yousaf, Muhammad Haroon
  • Velastin Carroza, Sergio Alejandro
In this study, a new multi-view human action recognition approach is proposed by exploiting low-dimensional motion information of actions. Before feature extraction, pre-processing steps are performed to remove noise from silhouettes, incurred due to imperfect, but realistic segmentation. Two-dimensional motion templates based on motion history image (MHI) are computed for each view/action video. Histograms of oriented gradients (HOGs) are used as an efficient description of the MHIs which are classified using nearest neighbor (NN) classifier. As compared with existing approaches, the proposed method has three advantages: (i) does not require a fixed number of cameras setup during training and testing stages hence missing camera-views can be tolerated, (ii) requires less memory and bandwidth requirements and hence (iii) is computationally efficient which makes it suitable for real-time action recognition. As far as the authors know, this is the first report of results on the MuHAVi-uncut dataset having a large number of action categories and a large set of camera-views with noisy silhouettes which can be used by future workers as a baseline to improve on. Experimentation results on multi-view with this dataset gives a high-accuracy rate of 95.4% using leave-one-sequence-out cross-validation technique and compares well to similar state-of-the-art approaches, Sergio A Velastin acknowledges the Chilean National Science and Technology Council (CONICYT) for its funding under grant CONICYT-Fondecyt Regular no. 1140209 (“OBSERVE”). He is currently funded by the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development and demonstration under grant agreement nº 600371, el Ministerio de Economía y Competitividad (COFUND2013-51509) and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Contribución de congreso (ConferenceOutput)

Characterisation of the spatial sensitivity of classifiers in pedestrian detection

  • Quinteros, Daniel
  • Velastin Carroza, Sergio Alejandro
  • Acuña Leiva, Gonzalo
This paper has been presented at : 6th LatinAmerican Conference on Networked Electronic Media (LACNEM-2015), In this paper, a study of the spatial sensitivity in the pedestrian detection context is carried out by a comparison of two descriptor-classifier combinations, using the well-known sliding window approach and looking for a well-tuned response of the detector. By well-tuned, we mean that multiple detections are minimised so as to facilitate the usual non-maximal suppression stage. So, to guide the evaluation we introduce the concept of spatial sensitivity so that a pedestrian detection algorithm with good spatial sensitivity can reduce the number of classifications in the pedestrian neighbourhood, ideally to one. To characterise spacial sensitivity we propose and use a new metric to measure it. Finally we carry out a statistical analysis (ANOVA) to validate the results obtained from the metric usage., The work described here was carried out as part of the OBSERVE project funded by the Fondecyt Regular programme of Conicyt (Chilean Research Council for Science and Technology) under grant no. 1140209. Sergio A Velastin has received funding from the Universidad Carlos III de Madrid, the European Unions Seventh Framework Programme for research, technological development and demonstration under grant agreement no 600371, el Ministerio de Economía y Competitividad (COFUND2013-51509) el Ministerio de Educación, cultura y Deporte (CEI-15-17) and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Evaluation of polyolefin-based macroporous separators for high temperature Li-ion batteries

  • Martínez Cisneros, Cynthia Susana
  • Levenfeld Laredo, Belén
  • Varez, Alejandro
Macroporous separators are critical components in liquid electrolyte batteries. Besides preventing physical contact between electrodes, they enable free ionic transport, electronic isolation and thermal shutdown. Nevertheless, separators also increase electrical resistance and takes up limited space inside the battery, affecting ionic conductivity. Widely used in lithium-ion batteries, commercial polyolefin-based separators operate in a limited temperature range, mainly ranging from -20 degrees C to + 60 degrees C. The purpose of this contribution is to assess the possibility to use these separators in lithium-ion batteries operating at extended temperatures, i.e. between -20 degrees C and 120 degrees C. For this purpose, four commercially available macroporous separators based on polyethylene and polypropylene, were investigated. To determine the effect of temperature on their performance, they were aged for one week at 120 degrees C. Evolution of their morphology and thermomechanical behavior was investigated using XRD, SEM, DSC, TGA and DMA. The thermal aging impact on the ionic conductivity was also investigated using LP30 (R) as reference electrolyte. Thermal aging, i.e. partial clogging of the porosity, was found to have significant effects mainly on mechanical strength, morphology and conductivity. (C) 2016 Elsevier Ltd. All rights reserved., This work has been supported by Projects funded by the regional government (Comunidad de Madrid through MATERYENER3CM S2013/MIT-2753) and the Spanish Government, MICINN (MAT2013-46452-C4-3R).J-Y Sanchez acknowledges the CONEX Programme, funding received from Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development and demonstration (Grant agreement n 600371), Spanish Ministry of Economy and Competitiveness (COFUND2013- 40258) and Banco Santander.
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Contribución de congreso (ConferenceOutput)

AI as a legal person

  • Laukyte, Migle
Abstract: The idea of the legal personhood of artificial intelligence (AI) — the idea that intelligent agents can have rights and incur obligations under the law— is controversial, and in fact is often dismissed out of hand: in this paper I will argue that, on the contrary, such legal personhood may be the next big challenge for our legal systems, and we need it to deal with the new kinds ofcomplexity introduced by AI. Furthermore, I argue that we already have experiences we can look: to this end we can draw on the reasoning applied to the legal personhood recognized for corporations and other nonhuman entities. In order to do this, I address some of the criticisms against ascribing legal personhood to AI. I also look at the Canadian and EU ethical guidelines so as to keep the development of AI within the framework of human values, and I show that an ascription of legal personhood to AI is consistent with them. I also address a few of the big issues involved in making the legal personhood of AI a reality., This paper is part of the project supported by the CONEX programme and has received funding from the Universidad Carlos III de Madrid, the European Union’s Seventh Framework Programme for research, technological development and demonstration under grant agreement N. 600371, el Ministerio de Economia, Industria y Competitividad (COFUND2014-51509), el Ministerio de Educacion, Cultura y Deporte (CEI-15-17) and Banco Santander. Universidad Carlos III de Madrid (APC. Read & Publish Agreement CRUE-CSIC 2023)
e-Archivo. Repositorio Institucional de la Universidad Carlos III de Madrid
Publication Artículo científico (JournalArticle)

Improving the efficiency of gas turbine systems with volumetric solar receivers

  • Petrakopoulou, Foteini Konstantina
  • Sánchez Delgado, Sergio
  • Marugán Cruz, Carolina
  • Santana Santana, Domingo José
The combustion process of gas turbine systems is typically associated with the highest thermodynamic inefficiencies among the system components. A method to increase the efficiency of a combustor and, consequently that of the gas turbine, is to increase the temperature of the entering combustion air. This measure reduces the consumption of fuel and improves the environmental performance of the turbine. This paper studies the incorporation of a volumetric solar receiver into existing gas turbines in order to increase the temperature of the inlet combustion air to 800 °C and 1000 °C. For the first time, detailed thermodynamic analyses involving both energy and exergy principles of both small-scale and large-scale hybrid (solar-combined cycle) power plants including volumetric receivers are realized. The plants are based on real gas turbine systems, the base operational characteristics of which are derived and reported in detail. It is found that the indications obtained from the energy and exergy analyses differ. The addition of the solar plant achieves an increase in the exergetic efficiency when the conversion of solar radiation into thermal energy (i.e., solar plant efficiency) is not accounted for in the definition of the overall plant efficiency. On the other hand, it is seen that it does not have a significant effect on the energy efficiency. Nevertheless, when the solar efficiency is included in the definition of the overall efficiency of the plants, the addition of the solar receiver always leads to an efficiency reduction. It is found that the exergy efficiency of the combustion chamber depends on the varying air-to-fuel ratio and, in most cases, it is maximized somewhere between the applied inlet combustion air temperatures of 800 °C and 1000 °C., F.P. would like to thank the Universidad Carlos III de Madrid, the European Union's Seventh Framework Programme for research, technological development and demonstration (grant agreements no 600371 and 332028), the Ministerio de Economía y Competitividad (COFUND2014-51509) the Ministerio de Educación, cultura y Deporte (CEI-15-17) and Banco Santander. S.S and C. would like to express their gratitude to the financial support of the Spanish Ministry of Economy and Competitiveness from project ENE2014-54942-R.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Long-term antipsychotic and benzodiazepine use and brain volume changes in schizophrenia: The Northern Finland Birth Cohort 1966 study

  • Huhtaniska, Sanna
  • Jaaskelainen, Erika
  • Heikka, Tuomas
  • Moilanen, Jani S.
  • Lehtiniemi, Heli
  • Tohka, Jussi
  • Coupe, Pierrick
  • Bjornholm, Lassi
  • Koponen, Hannu
  • Veijola, Juha
  • Isohanni, Matti
  • Kiviniemi, Vesa
  • Murray, Graham K.
  • Miettunen, Jouko
  • Manjón Herrera, José Vicente|||0000-0001-6640-927X
[EN] High doses of antipsychotics have been associated with loss in cortical and total gray matter in schizophrenia. However, previous imaging studies have not taken benzodiazepine use into account, in spite of evidence suggesting adverse effects such as cognitive impairment and increased mortality. In this Northern Finland Birth Cohort 1966 study, 69 controls and 38 individuals with schizophrenia underwent brain MRI at the ages of 34 and 43 years. At baseline, the average illness duration was over 10 years. Brain structures were delineated using an automated volumetry system, volBrain, and medication data on cumulative antipsychotic and benzodiazepine doses were collected using medical records and interviews. We used linear regression with intracranial volume and sex as covariates; illness severity was also taken into account. Though both medication doses associated to volumetric changes in subcortical structures, after adjusting for each other and the average PANSS total score, higher scan-interval antipsychotic dose associated only to volume increase in lateral ventricles and higher benzodiazepine dose associated with volume decrease in the caudate nucleus. To our knowledge, there are no previous studies reporting associations between benzodiazepine dose and brain structural changes. Further studies should focus on how these observations correspond to cognition and functioning., This study was supported by the University of Oulu Scholarship Foundation, the Jalmari and Rauha Ahokas foundation, the Orion Research Foundation sr, the Scholarship Fund of the University of Ouu - Tyyni Tani Found, the Foundation for Psychiatric Research, the Sigrid Juselius Foundation, the Academy of Finland (#132071, #268336 and #278286), the Brain & Behavior Research Foundation, Northern Finland Health Care Support Foundation, the Universidad Carlos III de Madrid, the European Union's Seventh Framework Programme for Research, Technological Development and Demonstration under Grant agreement no. 600371, el Ministerio de Economia y Competitividad (COFUND2013-40258), el Ministerio de Educacion, cultura y Deporte (CEI-15-17) and Banco Santander
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publication Artículo científico (JournalArticle)

Simulation and analysis of a stand-alone solar-wind and pumped-storage hydropower plant

  • Petrakopoulou, Fontina
  • Robinson, Alexander
  • Loizidou, Maria
This work presents the simulation and evaluation of a renewable hybrid power plant for off-grid fully autonomous operation on an intermediate-sized island in the Aegean Sea. A stand-alone energy system including storage facilities is simulated, optimized and analyzed relying on real-case weather and demand data of a relatively large remote community. Optimization of the power plant structure shows that to ensure continuous off-grid energy generation, even under extreme conditions, the combination of more than one renewable technology is required. The hybrid power plant consists of a pumped-storage hydropower plant, photovoltaic cells and wind turbines. Energy surplus of the power plant is used in the incorporated electrolyzer to generate a secondary product, hydrogen. Robust operation of the plant results in 48% of the energy generated stemming from the photovoltaic system and 52% from the wind turbines. The pumped-storage hydropower plant has a mean annual power output of 1.0 MW. The total mean annual efficiency of the hybrid plant is found to be 14.4%. Although stand-alone operation was achieved with the proposed plant, this requirement led to net energy output restrictions, capacity oversizing and large storage facilities., Fontina Petrakopoulou would like to thank the Universidad Carlos III de Madrid, the European Union's Seventh Framework Programme for research, technological development and demonstration (grant agreements nº 600371 and 332028), the Ministerio de Economía y Competitividad (COFUND2014-51509) and Banco Santander., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publication Artículo científico (JournalArticle)

Exergetic analysis and dynamic simulation of a solar-wind power plant with electricity storage and hydrogen generation

  • Petrakopoulou, Fontina
  • Robinson, Alexander
  • Loizidou, Maria
The ambitious vision of off-grid renewable energy autonomy of remote regions has yet to come to fruition. The development of comprehensive energy production systems would be needed to achieve such a goal. This study consists of the simulation and exergetic evaluation of a novel hybrid power plant for stand-alone operation aiming to provide electricity autonomy of a Mediterranean island. The considered power plant is simulated dynamically over an annual cycle and accounts for both energy input fluctuations and electricity surplus. The plant combines a photovoltaic array with wind turbines for energy input, coupled with electricity storage and a hydrogen generation facility to stabilize the power output of the plant. Unlike other similar studies, the energy system presented here relies on real-case weather and demand data of a relatively large remote community and is optimized to ensure continuous operation – even under extreme conditions. It is seen that this stand-alone hybrid power plant constitutes a robust and secure alternative to the current conventional energy situation; the combined renewable technologies succeed in complementing each other and offer stable performance throughout the year without the requirement of additional support by fossil fuels. The mean annual exergetic efficiency of the plant is found to be 17.9%, producing approximately 25,000 MWh of electricity per year, along with a secondary product (hydrogen) produced in the electrolyzer of the plant. Although this additional product is associated with additional investment cost, it offers the possibility to stabilize the power plant's performance and can be used as an additional source of financial income for the community., Fontina Petrakopoulou would like to thank the Universidad Carlos III de Madrid, the European Union's Seventh Framework Programme for Research, Technological Development and Demonstration (grant agreements n° 600371 and 332028), the Ministerio de Economía y Competitividad (COFUND2014-51509) and Banco Santander., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publication Artículo científico (JournalArticle)

Simulation and evaluation of a hybrid concentrating-solar and wind power plant for energy autonomy on islands

  • Petrakopoulou, Fontina
  • Robinson, Alexander
  • Loizidou, Maria
Renewable energy sources can offer isolated communities the opportunity to regulate their energy use in a manner that best suits their needs. This paper presents the simulation and thermodynamic evaluation of a stand-alone hybrid power plant exclusively using renewable energy sources and storage technologies for the energy autonomy of a Mediterranean island. The study assumes stand-alone dynamic operation and investigates the sustainable and robust energy independence of the community under consideration, a remote area not connected to a centralized electrical grid. The analysis shows that the evaluated hybrid concentrating solar-wind power plant is a reliable alternative for satisfying the fluctuating electricity demand of the island. The plant achieves stable and controlled autonomous performance using the complementary character of solar and wind energy, combined with energy storage., Fontina Petrakopoulou would like to thank the Universidad Carlos III de Madrid, the European Union's Seventh Framework Programme for research, technological development and demonstration (grant agreements nº 600371 and 332028), the Ministerio de Economía y Competitividad (COFUND2014-51509) and Banco Santander., Peer reviewed