Research Internship ProjectDevelopment of a prototype for embedded object detection with bio-inspired retinas on roboticplatforms Context The LEAT lab is leader of the national ANR project DeepSee in collaboration with Renault, Propheseeand 2 other labs in neuroscience (CERCO) and computer science (I3S). This project aims at exploring abio-inspired approach to develop energy-efficient solutions for image […]
Plus d'infoResearch Internship ProjectUnsupervised learning of robotic multimodal data Context LEAT lab has been working for several years on the design of bio-inspired neural models. One ofthem is inspired by the self-organization of the biological brain. This model named ReSOM has beenpreviously applied to the classification of multimodal data such as the representation of digits fromvisual, […]
Plus d'infoFrançois VerdierLaboratoire LEAT, francois.verdier@univ-cotedazur.fr Introduction In the field of intelligent objects, which are capable of retrieving a whole category ofinformation (such as the temperature of an aqueous solution, the pressure in oil pipes,identification badge numbers, reaction control in a nuclear power plant, etc.) andtransmitting it to a dedicated database via a wireless channel, we are […]
Plus d'infoResearch Internship ProjectDynamic and Adaptive Spiking Neural Network Solutions for Energy-Autonomous IoT nodes ContextAs the Internet of Things (IoT) continues to evolve, the integration of Artificial Intelligence (AI) withedge computing, i.e., Edge AI, emerges as a powerful synergy. This combination leverages MachineLearning (ML) algorithms to locally process sensor data, offering real-time intelligent decision-making.However, at this […]
Plus d'infoResearch Internship ProjectFPGA Based Parallel SNN Solution for Energy-Autonomous IoT nodes ContextAs the Internet of Things (IoT) continues to evolve, the integration of Artificial Intelligence (AI) withedge computing, i.e., Edge AI, emerges as a powerful synergy. This combination leverages MachineLearning (ML) algorithms to locally process sensor data, offering real-time intelligent decision-making.However, at this level of […]
Plus d'infoResearch Internship ProjectAn energy proportional neuromorphic solution for SpikeNets. ContextSpiking neural networks are considered as the third generation of neural networks and could thusreplace the conventional networks used in machine learning in order to reduce energy consumptionof AI, especially in Edge applications.But taking advantage of SNN needs to efficiently parallelize their execution onto multipleneuromorphic (event-based) […]
Plus d'infoResearch Internship ProjectEZBird – Training of embedded neural networks for bird song detection ContextThis internship subject takes place in the eBRAIN group of LEAT laboratory that works on EmbeddedBio-inspiRed Artificial Intelligence and Neuromorphic architectures.More specifically, it takes part a collaboration with CERN in Geneva in order to design, develop anddeploy a wireless sensor network to […]
Plus d'infoEnglish version available here. Partenaires : Equipe CMA du LEAT en collaboration avec la société NEXESSEncadrants : Aliou Diallo, Philippe Le Thuc, Robert StarajDomaine : RFID, antennes reconfigurablesDate de commencement : Février 2024Durée : 3 ansLieu : LEAT, Bât. Forum, Campus Sophia Tech, 930 route des colles, 06903 Sophia Antipolis, France Contexte : Les travaux […]
Plus d'infoA post-doctoral position is available at Université Côte d’Azur! « Online/Incremental Learning for Spatial Applications » [Please find the full description in the pdf enclosed] Keywords: Artificial Intelligence, Online learning, Incremental/Continual Learning, Influence functions, data pruning, federated learning Mission: Executing machine learning (ML) algorithms on satellite could offer many advantages. It will indeed reduce the required bandwidth, […]
Plus d'infoLe modèle conceptuel de base pour une antenne d’émission VLF-LF est un monopôle électrique vertical placé sur un plan de sol parfaitement conducteur. Pour des antennes d’émission fonctionnant encore plus bas en fréquence, du fait de longueurs d’onde extrêmement importantes, les structures couramment utilisées sont horizontales et utilisent les caractéristiques de l’environnement pour optimiser leur rendement. Ces modèles d’antenne sont dictés par des raisons techniques et financières afin d’obtenir le meilleurs compromis coût – efficacité de rayonnement. La modélisation de telles structures constitue un défi pour les méthodes numériques par la prise en compte précise des caractéristiques de l’environnement nécessaire à l’évaluation des rendements.
Plus d'info