Réunion

Caméras événementielles et modèles neuromorphiques – Applications en robotique (Event-Based Vision and Neuromorphic Computing for Robotics)

Annonce

Les inscriptions réalisées entre le 17 juin et le 13 septembre n’ont pas été prises en compte; il est nécessaire de vous inscrire à nouveau à la journée.

Les demandes de prise en charge de mission par le GdR, sont acceptées jusqu’au 22 septembre.

Cette journée est co-organisée avec le GdR Robotique.

Objectifs de la journée / Scope and objectives

Thématiques d’intérêt / Topics of interest

Orateur.ice.s invité.e.s / Keynote speakers

Giulia D’Angelo (Assistant Professor, CTU Prague),

Presentation title: « What’s catching your eye? – Event-driven sensing and neuromorphic computing for active vision »

Abstract: Vision is an exploratory behaviour that emerges from the dynamic relationship between actions and sensory feedback. For any agent, whether biological or robotic, processing visual input efficiently is fundamental to understanding and interacting with the environment. The central challenge lies in continuously recalibrating perception through sensorimotor contingencies, where what an agent sees is shaped by how it moves. Embodiment is central to this vision: perception and action are inseparable, and intelligence emerges from their continuous interaction with the physical world, shaped by the very structure of our sensors. Selective visual attention is one of the many mechanisms by which the visual cortex meets this challenge, organizing and interpreting complex visual scenes in real time. To address these challenges, I develop brain-inspired algorithms that harness the computational principles of biological neurons and spiking neural networks, optimized for neuromorphic hardware. These algorithms enable real-time robotic perception with microsecond latency and milliwatt power consumption, bringing the efficiency of biological vision systems within reach of autonomous systems at the edge.

Bio: Giulia D’Angelo is an Assistant Professor at the Czech Technical University in Prague, where she develops neuromorphic algorithms for active vision. She earned a BSc in Biomedical Engineering from the University of Genoa and an MSc (with honours) in Neuroengineering. During her Master’s at King’s College London, she developed a neuromorphic system for the egocentric representation of peripersonal visual space. She completed her PhD in neuromorphic algorithms at the University of Manchester, where she received the President’s Doctoral Scholar Award, in collaboration with the Event-Driven Perception for Robotics Laboratory at the Italian Institute of Technology, proposing a biologically plausible model for event-driven, saliency-based visual attention. Following her PhD, she was awarded a Marie Skłodowska-Curie Postdoctoral Fellowship at the Czech Technical University in Prague, during which she explored sensorimotor contingency theories in neuromorphic active vision. After completing the fellowship, she joined the Czech Technical University in Prague as an Assistant Professor. Her current research, funded by the GACR Standard Grant from the Czech Science Foundation, bridges bio-inspired software and neuromorphic hardware to enable robust, efficient perception and control for low-power, low-latency autonomous systems.

Guillermo Gallego (Full Professor, TU Berlin),

Presentation title: « Seeking robustness in supervised and unsupervised low-level event-based vision »

Abstract: This talk will present recent progress (last two years) in advancing the solution of low-level vision task with event cameras. It will present both supervised and unsupervised approaches for depth, motion, intensity and noise estimation that generalize beyond the training dataset.

Bio: Guillermo Gallego is Full Professor at Technical University of Berlin and at the Einstein Center Digital Future, where he leads the Robotic Interactive Perception Laboratory. He is also a Principal Investigator at the Science of Intelligence Excellence Cluster and the Robotics Institute Germany. He received the Ph.D. degree in Electrical and Computer Engineering from the Georgia Institute of Technology, USA, in 2011, supported by a Fulbright Scholarship, and followed by a Marie Curie fellowship. He serves as Associate Editor for IEEE T-PAMI, RA-L and IJRR, and guest Editor for IEEE T-RO.

Benoît R. Cottereau (CNRS Research Director, CerCO, Toulouse)

Presentation title: « Robust Scene Understanding with Bio-Inspired and Efficient AI: From Event-Based Sensing to Spiking Neural Computation »

Abstract: Deep neural networks have achieved remarkable performance across a wide range of visual tasks, yet their reliance on dense visual representations and large numbers of real- valued parameters remains challenging for robust and efficient deployment in real-world and embedded settings. In this talk, I will explore how principles inspired by biological vision and neural computation can be leveraged to address these challenges by combining event-based sensing with spiking neural networks (SNNs). I will first discuss how the asynchronous and sparse representation of visual information provided by event-based cameras can be naturally exploited by spike-based computation for efficient visual scene understanding. I will then present recent work from my group on event-based visual place recognition, where a compact spiking architecture achieves competitive performance under substantial changes in viewpoint and illumination while drastically reducing model size and estimated energy consumption. Finally, I will turn to the temporal dynamics of SNNs themselves, showing how learnable delays in recurrent spiking networks can exploit temporal structure as a computational resource, drawing inspiration from the temporal dynamics of biological neural systems. Together, these studies illustrate how bio-inspired sensing and computation can provide new strategies for building robust, compact, and energy-efficient artificial vision systems.

Bio: Benoit R Cottereau is a CNRS research director at IPAL (IRL 2955, Singapore) and CerCo (UMR 5549, Toulouse, France) laboratories. He received the M. Sc. Degree in signal processing from Supelec (Paris Saclay, France) in 2004 and a Ph.D. degree in computational neurosciences from Paris Saclay University in 2008 before to complete a post-doctoral fellowship at Stanford University (USA). Benoit now leads the SV3M (‘Spatial Vision in Man, Monkey and Machine’) group at CerCo laboratory (UMR 5549, Toulouse, France) and is in charge of the efficient AI research program at IPAL (IRL 2955, Singapore). His research interests include visual processing in biological and artificial systems with a focus on visual scene understanding, e.g., semantic segmentation, depth estimation, and motion prediction.

Appel à contributions / Submission guidelines

Lieu / Address

Amphithéâtre Durand, campus Pierre et Marie Curie, Sorbonne Université, 4, place Jussieu, 75005 Paris (entrée place Jussieu, métro 7 et 10, puis suivre bâtiment Esclangon / entrance on Place Jussieu, Metro lines 7 and 10, then follow signs to the Esclangon building)

Dates importantes / Important dates

• 13 sept 2026 : Date limite d’envoi des résumés / Abstract submission deadline
• 15 sept 2026 : Notification aux auteurs / Author notification
• 6 oct 2026 : journée « Caméras événementielles et modèles neuromorphiques – Applications en robotique » / worshop « Event-Based Vision and Neuromorphic Computing for Robotics »

Organisateurs / Organisers

• Julien Moreau (Heudiasyc, Compiègne : julien.moreau@hds.utc.fr)
• Jean Martinet (I3S, Sophia-Antipolis : jean.martinet@univ-cotedazur.fr)
• Franck Davoine (LIRIS, Lyon : franck.davoine@cnrs.fr)

Programme / Schedule

Time	Program
09:30-10:00	Welcome coffee
10:00-10:10	Welcome and introduction
10:10-10:50	Keynote 1 — Benoit R. Cottereau (CerCO, Toulouse) — Robust Scene Understanding with Bio-Inspired and Efficient AI: From Event-Based Sensing to Spiking Neural Computation
10:50-11:35	Oral Session 1 — Spiking Neural Networks for Event-Based Vision
10:50-11:05	Amélie Gruel, Pierre Lewden, Adrien F. Vincent, Sylvain Saïghi — Line-based Event Preprocessing: Towards Low-Energy Neuromorphic Computer Vision
11:05-11:20	Wasi Ullah, Sébastien Ambellouis, Charles Tatkeu — Spiking Transformer Framework for Event-Based Object Detection
11:20-11:35	Alimatou Sadia Memudu, Jean Martinet, Yuta Nakano — MAG-Voxel: Motion-Aware Event Representation for Efficient SNN-Based Person Detection
11:35-12:15	Keynote 2 — Guillermo Gallego (TU Berlin) — Seeking robustness in supervised and unsupervised low-level event-based vision
12:15-13:45	Lunch break (numerous lunch options in the neighborhood)
13:45-14:30	Oral Session 2 — Neuromorphic and Efficient Event-Based Computing
13:45-14:00	Gnouyadou Romaric Mazna, Sai Deepesh Pokala, Jean Martinet — Exploring deep learning for event-based saliency prediction with a transformer-based model.
14:00-14:15	Doha Benjelloun, David Roussel, Fabien Bonardi, Samia Bouchafa — Directional Selective Filters for Guiding Spiking Neural Networks in Event-Based Optical Flow Estimation
14:15-14:30	Sasskia Brüers, Gilles Bézard, Douglas McLelland — Standard CNNs, Event-Driven Hardware: Exploiting Activation Sparsity for Sub-Millisecond Inference
14:30-15:15	Poster session & coffee break
15:15-15:55	Keynote 3 — Giulia D’Angelo (CTU Prague) — What’s catching your eye? – Event-driven sensing and neuromorphic computing for active vision
15:55-16:25	Oral Session 3 — Event-Based Scene and Motion Understanding
15:55-16:10	Elisa Lannelongue, Cédric Demonceaux, Guillaume Caron, Carlos Mateo-Agulló, Fumio Kanehiro — Event-based 6-DOF motion estimation under depth uncertainty
16:10-16:25	Abdessamad El Kaouri, Mohamed Kas, Yassine Ruichek, Youssef El Merabet — Synthetic-to-Real Event Domain Adaptation for Semantic Segmentation in Driving Scenes
16:25-16:30	Closing remarks and discussion

• Laure Acin, Pierre Jacob, Camille Simon-Chane, Aymeric Histace — ParallelizedGRU: a Generic Memory-Efficient Recurrent Architecture.
• Gnouyadou Romaric Mazna, Sai Deepesh Pokala, Jean Martinet — Human gaze-based filtering for action recognition.
• Reda Remiten, Jonathan Ledy, Benoit Vigne — Low-cost embedded architecture for online object detection using an event camera.
• Kévin Hoarau, Louis Airale, Stefan Duffner, Franck Davoine, Fadi Dornaika — An attention-based mechanism into hierarchical graph architectures for event-based cameras.
• Pedro Sacramento Xavier Barreto Rosa, Jean Martinet, Yuta Nakano — Multimodal SNNs and the Diverging Path Ahead.
• Ivan Gutierrez Rodriguez, Julien Moreau, Chiara Bartolozzi, Arren Glover — E-MOTION: A Dataset for Event-Based Scene Flow Estimation with Independent Moving Objects.
• Paul Longour, Julien Moreau, Franck Davoine — Bringing BNNs to Fast Event Processing.
• Geoffroy Keime, Nicolas Cuperlier, Benoit R. Cottereau — REACT: A Fully Spiking State-Space Model for Real-Time Event-Driven Temporal Perception
• David Dirnfeld, Peter Xie, Erik Learned-Miller — Image acquisition with events cameras via blinking
• Hugo Bulzomi, Alimatou Sadia Memudu, Crescenzo Edoardo Mauriello, Rémy Bendahan, Takeshi Fujita, Yuta Nakano, Jean Martinet — Towards Low-Energy Neuromorphic Computer Vision

Remerciements : Cette journée est co-financée par le GdR IASIS, le GdR Robotique, et le projet ANR REVE-BNN (ANR-24-CE33-4001), et bénéficie de l’aimable soutien du SCAI (Sorbonne Cluster for Artificial Intelligence).