fg2026

Tutorial: Face and gesture recognition using event-based cameras

FG2026 Tutorial, Kyoto, Japan

Tutorial description

Event-based cameras are a recently introduced technology that asynchronously senses light intensity changes at the individual pixel level. They are increasingly being applied to various recognition problems of interest to the computer vision community, primarily because they continuously encode sparse appearance and motion information at very high speeds, with low latency, and across a high dynamic range. Over the last few years, interest in these devices has surged, particularly regarding their application to face and gesture analysis. Event-based cameras are uniquely suited for these tasks due to their high temporal resolution and high dynamic range, which eliminate motion blur and accommodate challenging illumination conditions. Moreover, they enable the analysis of subtle facial changes within a continuous data stream. In this context, this tutorial provides an introductory yet comprehensive overview of event-based cameras and discusses their application to face and gesture recognition. The tutorial is organized into three parts. First, we provide an overview of event cameras, existing sensors, and the problems to which they have been applied, ranging from low-level vision (e.g., optical flow, tracking, and feature detection) to high-level vision (e.g., reconstruction, segmentation, and recognition). Second, we discuss existing techniques for processing event streams, including learning-based approaches. Finally, we present an overview of recent work on face and gesture recognition using event-based cameras, including a discussion of available datasets, current methodologies, and open research directions. This tutorial is intended both for researchers with no prior experience with event-based cameras and for those who have worked with them in the past but wish to review recent advances in the field.

Schedule:

13:30-15:00 Opening and Basics of Event-based cameras

15:00-16:10 Event-based cameras for Face and Gesture Recognition

16:10-16:20 Future Perspectives and Closing Remarks

Tutorial Slides:

Requirements (e.g., facilities, Internet access, etc.) and Additional Information:

Organizers / Lecturers:

Rodrigo Verschae

Daniel Acevedo

Nicolas Mastropasqua:

Ignacio Bugueno-Cordova:

Experience

The instructors possess extensive experience in face and gesture detection and recognition problems using event-based cameras [1–5] (see the list of related publications at the end of this document). They also have a strong background in facial recognition, detection, and analysis using standard RGB cameras. Furthermore, the instructors have previously delivered tutorials and talks on this topic, including a tutorial titled “Introduction to Face and Gesture Recognition Using Event-Based Cameras” at the 15th IEEE International Conference on Automatic Face and Gesture Recognition (FG 2020). This updated tutorial builds upon the 2020 edition, incorporating the substantial body of work that has been published over the last few years.

FG + Event-based Related Publications authored by the lectures