Event-Based Vision and Neuromorphic Computing for Robotics

Event-based cameras are emerging as a compelling alternative to conventional frame-based imaging sensors. By producing asynchronous and sparse visual information with very high temporal resolution and dynamic range, they are particularly well suited to robotic systems operating under fast dynamics, real-time constraints, and strict limitations in computation and energy consumption.

At the same time, the development of dedicated representations and learning paradigms, ranging from convolutional and Transformer-based architectures to spiking neural networks (SNNs), has significantly advanced the processing of event-based data. In particular, SNNs and neuromorphic approaches are highly attractive for embedded and robotic applications due to their event-driven nature and their compatibility with neuromorphic hardware, enabling low-latency, energy-efficient, and online computation.

The workshop aims to bring together the research community in France working on event-based perception for robotics, while remaining open to international contributions. The workshop will foster interdisciplinary exchanges between researchers in signal processing, computer vision, machine learning, neuromorphic computing, and robotics.

The program will include three invited talks by leading French and international researchers, as well as contributed oral presentations.

Topic of interest 

We invite submissions on recent or ongoing research, both methodological and application-oriented, including but not limited to:

• Event-based sensors and neuromorphic cameras for robotics
• Representation and processing of event streams
• Efficient models for event-based data processing
• Machine learning for event-based vision
  • Spiking Neural Networks (SNNs)
  • CNNs, RNNs, GNNs, SSMs, and Transformers applied to event-based data
• Neuromorphic learning and computing
• Dedicated hardware architectures and embedded neuromorphic systems
• Robotic applications: perception, navigation, SLAM, interaction, manipulation…

Keynote speakers

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

Benoît Cottereau (CNRS Research Director, CerCO, Toulouse)

Guillermo Gallego (Full Professor, TU Berlin),

Submission guidelines

Contributions from doctoral students and early-career researchers are encouraged.

Submission of a one or two-page abstract is required (title, authors, institution, and summary). Will be selected for either an oral presentation or a poster.
Please send your proposals by e-mail to all the organisers: J. Moreau ([email protected]), J. Martinet ([email protected]), F. Davoine ([email protected]).

All presentations will be made in English.