Pasar al contenido principal

COOPERACION DE VEHICULOS DE SUPERFICIE Y AEREOS PARA APLICACIONES DE INSPECCION EN ENTORNOS CAMBIANTES

PID2021-127648OB-C33

•
Nombre agencia financiadora Agencia Estatal de Investigación
Acrónimo agencia financiadora AEI
Programa Programa Estatal para Impulsar la Investigación Científico-Técnica y su Transferencia
Subprograma Subprograma Estatal de Generación de Conocimiento
Convocatoria Proyectos de I+D+i (Generación de Conocimiento y Retos Investigación)
Año convocatoria 2021
Plan Estatal 2021-2023 Plan Estatal de Investigación Científica y Técnica y de Innovación
Centro beneficiario UNIVERSIDAD COMPLUTENSE DE MADRID

Resultados relacionados

Resultados totales (Incluyendo duplicados): 7
Encontrada(s) 1 página(s)

Docta Complutense
Publicación Artículo científico (JournalArticle)

Customization of the text-to-image diffusion model by fine-tuning for the generation of synthetic images of cyanobacterial blooms in lentic water bodies

  • Barrientos-Espillco, Fredy
  • Pajares Martínsanz, Gonzalo
  • López Orozco, José Antonio
  • Besada Portas, Eva
Cyanobacterial blooms emerge unpredictably on the surface of lentic water bodies, posing both ecological threats and public health risks. To effectively monitor these events, this study introduces the use of Machine Vision Systems (MVS) integrated into Autonomous Surface Vehicles (ASVs). These ASVs are capable of autonomous and safe navigation, enabling them to detect cyanobacterial blooms while avoiding obstacles. Convolutional Neural Networks (CNNs) are employed for early detection and continuous monitoring, but their effectiveness hinges on access to large, high-quality training datasets. Due to the sporadic and uncontrollable nature of bloom occurrences, acquiring sufficient real-world images for training and validating CNN models is a significant challenge. To overcome this, the Stable Diffusion XL (SDXL) text-to-image generative model is utilized to produce realistic synthetic images, ensuring a sufficient dataset for training. However, SDXL alone struggles to accurately depict cyanobacterial blooms. To address this limitation, DreamBooth is used to fine-tune SDXL with a small set of real bloom-specific image patches. To ensure the diversity of the synthetic dataset, detailed prompts for SDXL are generated using a Large Language Model (LLM). The combination of SDXL fine-tuning with LLM-driven prompts design applied to environmental monitoring and autonomous navigation in lentic environments represents the core innovation of this work. A dual-task CNN model is then trained on the synthetic dataset to simultaneously detect blooms and obstacles. Experimental results demonstrate the effectiveness and novelty of the proposed approach, showing improvements of up to 15.74% in object detection and 6.48% in semantic segmentation compared to the baseline dataset.
Docta Complutense
Publicación Artículo científico (JournalArticle)

Analyzing and interpreting convolutional neural networks using latent space topology

  • López González, Clara Isabel
  • Gómez Silva, María José
  • Besada Portas, Eva
  • Pajares Martínsanz, Gonzalo
The development of explainability methods for Convolutional Neural Networks (CNNs), under the growing framework of explainable Artificial Intelligence (xAI) for image understanding, is crucial due to neural networks success in contrast with their black box nature. However, usual methods focus on image visualizations and are inadequate to analyze the encoded contextual information (that captures the spatial dependencies of pixels considering their neighbors), as well as to explain the evolution of learning across layers without degrading the information. To address the latter, this paper presents a novel explanatory method based on the study of the latent representations of CNNs through their topology, and supported by Topological Data Analysis (TDA). For each activation layer after a convolution, the pixel values of the activation maps along the channels are considered latent space points. The persistent homology of this data is summarized via persistence landscapes, called Latent Landscapes. This provides a global view of how contextual information is being encoded, its variety and evolution, and allows for statistical analysis. The applicability and effectiveness of our approach is demonstrated by experiments conducted with CNNs trained on distinct datasets: (1) two U-Net segmentation models on RGB and pseudo-multiband images (generated by considering vegetation indices) from the agricultural benchmark CRBD were evaluated, in order to explain the difference in performance; and (2) a VGG-16 classification network on CIFAR-10 (RGB) was analyzed, showing how the information evolves within the network. Moreover, comparisons with state-of-the-art methods (Grad-CAM and occlusion) prove the consistency and validity of our proposal. It offers novel insights into the decision making process and helps to compare how models learn.
Docta Complutense
Publicación Artículo científico (JournalArticle)

Enseñanza de robótica móvil en el Robotarium-UCM

  • Villar Galán, Diego
  • García Pérez, Lía
  • Chacón Sombría, Jesús
Este artículo presenta una extensión de la plataforma Robotarium-UCM orientada a la enseñanza de robótica multiagente, mediante la integración de un sistema de comunicaciones distribuido basado en la librería 0MQ. La arquitectura propuesta se fundamenta en un servidor central que actúa como un conmutador lógico, permitiendo definir topologías de comunicación entre agentes a través de grafos dirigidos. Estas estructuras no solo permiten implementar interacciones flexibles y realistas entre robots, sino que reflejan directamente los modelos usados en teoría de consenso y control multiagente, facilitando el análisis de estabilidad mediante operadores como el laplaciano. La implementación de esta arquitectura ha permitido diseñar prácticas docentes más cercanas al estado del arte en investigación y desarrollo de sistemas cooperativos.
Docta Complutense
Publicación Artículo científico (JournalArticle)

Introduction to Control Education with the Unibotics web framework

  • García Pérez, Lía
  • Martín-Martín, Diego
  • Cañas Plaza, José María
  • Chacón Sombría, Jesús
  • Roldán Álvarez, David
This paper presents the Unibotics platform, a simulation framework for teaching robotics in university courses. It allows students to program robots from the web browser in Python, using novel and current technologies such as ROS middleware and Gazebo robotics simulator. Unibotics provides an interface that hides the initial complexity of installing and dealing directly with ROS, Gazebo or MoveIt and allows students to focus on the algorithmic aspects of robotics. In addition, this article describes two teaching experiences of using Unibotics with real higher education students to introduce robot control. First, with Computer Science students in an elective Robotics course at U. Complutense Madrid. Second, with students of "Automation and Industrial Robotics" course of Industrial Electronics and Automation Engineering degree at U. Rey Juan Carlos. Both experiences have been successful and they will be continued in the next course.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Design, Control and Real-World Deployment of an Autonomous Surface Vehicle for Water Quality Monitoring, Diseño, control y despliegue en entorno real de un vehículo autónomo de superficie para la monitorización de la calidad del agua

  • Rochas , Alejandro
  • Forero , David
  • González , Armando
  • García Pérez, Lía
  • Chacón Sombría, Jesús
  • Lopez-Orozco, Jose Antonio
[ES] La monitorización de cianobacterias en embalses representa un desafío crítico para la gestión sostenible de recursos hídricos. Este artículo presenta el diseño, control y despliegue en un entorno real de un vehículo autónomo de superficie (VAS) compacto y portátil, equipado con un cabestrante y sonda multiparamétrica capaz de realizar perfiles verticales de forma automatizada para la obtención de parámetros de la calidad del agua. Se detalla su arquitectura hardware/software, control, estrategias de navegación y planificación de misiones, así como los resultados del despliegue en el embalse de Guajaraz, donde demostró su capacidad para operar de forma autónoma, obteniendo datos de alta resolución espacial y temporal con un coste operativo reducido. Esta plataforma representa una solución escalable para programas de monitorización continua en aguas continentales.
Docta Complutense
Publicación Artículo científico (JournalArticle)

Layer factor analysis in convolutional neural networks for explainability

  • López González, Clara Isabel
  • Gómez Silva, María José
  • Besada Portas, Eva
  • Pajares Martínsanz, Gonzalo
Explanatory methods that focus on the analysis of the features encoded by Convolutional Neural Networks (CNNs) are of great interest, since they help to understand the underlying process hidden behind the black-box nature of these models. However, to explain the knowledge gathered in a given layer, they must decide which of the numerous filters to study, further assuming that each of them corresponds to a single feature. This, coupled with the redundancy of information, makes it difficult to ensure that the relevant characteristics are being analyzed. The above represents an important challenge and defines the aim and scope of our proposal. In this paper we present a novel method, named Explainable Layer Factor Analysis for CNNs (ELFA-CNNs), which models and describes with quality convolutional layers relying on factor analysis. Regarding contributions, ELFA obtains the essential underlying features, together with their correlation with the original filters, providing an accurate and well-founded summary. Through the factorial parameters we gain insights about the information learned, the connections between channels, and the redundancy of the layer, among others. To provide visual explanations in a similarly way to other methods, two additional proposals are made: a) Essential Feature Attribution Maps (EFAM) and b) intrinsic features inversion. The results prove the effectiveness of the developed general methods. They are evaluated in different CNNs (VGG-16, ResNet-50, and DeepLabv3+) on generic datasets (CIFAR-10, imagenette, and CamVid). We demonstrate that convolutional layers adequately fit a factorial model thanks to the new metrics presented for factor and fitting residuals (D1, D>, and Res, derive from covariance matrices). Moreover, knowledge about the deep image representations and the learning process is acquired, as well as reliable heat maps highlighting regions where essential features are located. This study effectively provides an explainable approach that can be applied to different CNNs and over different datasets.
Docta Complutense
Publicación Artículo científico (JournalArticle)

Vehículo autónomo de superficie para la monitorización de cianobacterias en embalses. Un caso práctico

  • Rochas Fernández, Alejandro
  • Forero Martínez, David Eugenio
  • González García, Armando David
  • Chacón Sombría, Jesús
  • García Pérez, Lía
  • López Orozco, José Antonio
Ante el aumento de la población y el cambio climático, los recursos hídricos disponibles están siendo amenazados. Las autoridades de los países avanzados buscan nuevos modelos y sistemas de vigilancia que garanticen la calidad del agua frente a los nuevos desafíos y contaminantes emergentes. Sin embargo, los métodos actuales de monitorización no son aplicables, tanto por coste como por inmediatez ante el aumento que se requiere en la monitorización tanto espacial como en frecuencia. El uso de vehículos autónomos de superficie (VAS) es una solución, escalable y económica, que permite una mayor frecuencia de monitorización en cualquier parte un embalse a un coste efectivo bajo. Este artículo describe el VAS desarrollado en la UCM para la monitorización de cianobacterias, contaminante biológico emergente; el procedimiento para desplegarlo en un embalse real; y, que es fácil aumentar el número de muestreos con un coste mucho menor que los métodos actuales. Sin embargo, aunque el vehículo es autónomo, se necesita la intervención humana.