Pasar al contenido principal

Resultados relacionados

Resultados totales (Incluyendo duplicados): 3
Encontrada(s) 1 página(s)

DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

A 128×128 Electronically Multi-Foveated Dynamic Vision Sensor With Real-Time Resolution Reconfiguration

  • Faramarzi, Farnaz
  • Linares-Barranco, Bernabé
  • Serrano-Gotarredona, Teresa
This paper presents the design and implementation of a 128× 128 electronically foveated dynamic vision sensor (EF-DVS) fabricated using 350 nm CMOS technology. The EF-DVS integrates a novel pixel grouping approach that permits real-time dynamic resolution adjustments via external digital signals. Previous approaches rely on physically crafting high- and low-resolution regions, which require a mechanical setup for tracking moving objects within the fovea. Here, our innovation supports flexible and fast operation modes, acting on amplified photocurrents, allowing the sensor to operate in both high-resolution and low-resolution settings, and to configure multiple high-resolution regions of interest (ROIs) with arbitrary shapes and sizes within the pixel array in real-time. Although the pixel circuitry is more complex than its un-foveated predecessor, we have kept the same pixel area, sacrificing slightly fixed pattern noise (FPN). The sensor achieves a latency of 3.66μs and demonstrates a contrast sensitivity down to 1.03%. It maintains a dynamic range exceeding 120 dB and an intra-scene dynamic range above 60 dB. Notably, the power consumption is reduced with respect to its predecessor, down to 0.7 mW at 100 Keps when configured in low-resolution mode with a 2× 2 pixel grouping. The ability to dynamically adjust spatial resolution reduces noise event rates, enhances sensitivity, and lowers both data bandwidth and processing requirements. These features make the EF-DVS a suitable candidate for applications in robotics, surveillance, and real-time monitoring systems where efficient data processing and low latency are critical., This work was supported in part by EU CHIST-ERA Grants APROVIS3D under Project PCI2019-111826-2; in part by EU HE NIMBLEAI under Grant 101070679; in part by the Andalusian Regional Grant BIOVEO under Project PROYEXCEL_00060; in part by Spanish Grants from the Ministry of Science, Innovation and Universities MULTI-SENS under Grant PDC2023-145841-C31; and in part by the NANOMIND through European Regional Development Fund under Grant PID2019-105556GB-C31. The work of Farnaz Faramarzi was supported by the FPI Scholarship from the Spanish Ministry of Science, Innovation, and Universities., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

BAM-SLDK: biologically inspired attention mechanism with spiking learnable delayed kernel synapses

  • Chacón-Falcón, Mario
  • Patiño-Saucedo, Alberto
  • Camuñas-Mesa, Luis A.
  • Serrano-Gotarredona, Teresa
  • Linares-Barranco, Bernabé
Spiking neural networks are emerging as an alternative neural network model due to their biological plausibility, energy efficiency, and built-in ability to learn from temporal dynamics. However, in order to effectively process data with rich spatial and temporal dependencies, the usual static projections (feedforward and recurrent) among layers of spiking neurons fail to represent all the information needed. Inspired by how synaptic delays affect the learning process in biological neurons, in this paper, we propose a biologically inspired attention mechanism based on spiking convolutions with learnable delayed kernel synapses. The proposed model increases temporal learning ability, attending simultaneously to spatial and temporal dynamics with few parameters required. More precisely, our main technical contributions are: (1) we add kernels to the temporal dimension to enlarge the receptive field of the convolution; (2) we time kernels activations to mimic multiple delayed times; and (3) we introduce three different pruning techniques to optimize the number of delays and parameters used. Experiments show that our method surpasses conventional spiking convolutional modules and achieves state-of-the-art results. When pruning, we show that, for some datasets or pruning techniques, removing up to 80% of the initially trained delays results in minimal performance loss, effectively reducing memory consumption and parameters required. To the best of our knowledge, this is the first time that learnable delayed synapses have been included in spiking convolutional layers for neuromorphic datasets classification, unlocking a new biologically inspired attention mechanism and achieving superior performance on high temporal demanding tasks., This work was funded by the Ministry for Digital Transformation and Public Function through grant USECHIP (TSI-069100-2023-001) of PERTE Chip Chair program, funded by European Union – Next Generation EU, and partly funded by Grant SAMURAI PDC2023-145841-C31 from the Spanish Ministry of Science, Innovation, and Universities, and by regional Grant BIOVEO PROYEXCEL_00060. M C F was supported by CSIC JAE-Intro-ICU fellowship AIHUB-15., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

Neuromorphic Computing and Applications: A Topical Review

  • Enuganti, Pavan Kumar
  • Sen-Bhattacharya, Basabdatta
  • Serrano-Gotarredona, Teresa
  • Rhodes, Oliver
Neuromorphic computers achieve energy efficiency by emulating brain structure and event-driven processing that reduces energy consumption significantly. An increasing interest in this technology started in the initial years of this millennium, sparked by the awareness and concern on the ever-increasing power demands of modern-day computing. In current times, there are several neuromorphic computers and sensors that continue to be developed in both industry and academic research. The focus of this survey is on the neuromorphic computing applications of these devices that include brain-inspired neural networks, brain-inspired artificial neural networks, and Hybrid circuits comprising both artificial and brain-inspired units of computation. Many of these applications use neuromorphic sensors as input devices. We have surveyed three specific neuromorphic computers viz. SpiNNaker, TrueNorth, Loihi, and one neuromorphic sensor viz. Dynamic vision sensor (DVS)-based electronic retina; the demonstration of neuromorphic computing and applications using these devices far outnumbers those on the others that are currently available, which forms the basis of our choice. The applications include low-power cognitive machine intelligence as well as neuropathological understanding and knowledge discovery. Overall, our survey identifies the potential for neuromorphic computing to provide low power, low cost, and dynamic solutions for societal and scientific problems in the not-too-distant future., EPK was supported by the BITS Pilani Goa Campus Institute Fellowship awarded towards his Doctoral Research. Part of this research was supported by the Science and Engineering Research Board (SERB) Core Research Grant CRG/2019/003534 to BSB. TSG was funded by Spanish National Projects MEMVIS (PDC2023-145841-C31 / AEI / 10.13039/501100011033 and the EU PRTR funds-NextGenerationEU) and EUPHORIC (PID2023-149071NB-C51 funded by MCIU/ AEI / 10.13039/501100011033/FEDER,UE). OR was supported in part through the NimbleAI project, which has received funding from the EU's Horizon Europe Research and Innovation programme (Grant Agreement 101070679), and by UK Research and Innovation (UKRI) under the UK government's Horizon Europe funding guarantee (Grant Agreement 10039070). The research is also supported by the EPSRC UK through the Edgy Organism research project (Grant agreement EP/Y030133/1)., Peer reviewed