EV-Gait: Event-Based Robust Gait Recognition Using Dynamic Vision Sensors

TLDR

Event cameras offer ultra‑low power consumption, high temporal resolution, and a wide dynamic range, but they output noisy, asynchronous intensity change events instead of frames, limiting the use of conventional vision‑based gait recognition algorithms. The authors aim to develop EV‑Gait, an event‑based gait recognition method that uses motion consistency to denoise event streams and a deep neural network to classify gait. EV‑Gait denoises event streams via motion consistency and classifies gait with a deep neural network, evaluated on two datasets—one real‑world and one derived from the CASIA‑B RGB benchmark. Experiments demonstrate that EV‑Gait achieves about 96 % accuracy in real‑world settings and matches state‑of‑the‑art RGB‑based methods on the CASIA‑B benchmark.

Abstract

In this paper, we introduce a new type of sensing modality, the Dynamic Vision Sensors (Event Cameras), for the task of gait recognition. Compared with the traditional RGB sensors, the event cameras have many unique advantages such as ultra low resources consumption, high temporal resolution and much larger dynamic range. However, those cameras only produce noisy and asynchronous events of intensity changes rather than frames, where conventional vision-based gait recognition algorithms can't be directly applied. To address this, we propose a new Event-based Gait Recognition (EV-Gait) approach, which exploits motion consistency to effectively remove noise, and uses a deep neural network to recognise gait from the event streams. To evaluate the performance of EV-Gait, we collect two event-based gait datasets, one from real-world experiments and the other by converting the publicly available RGB gait recognition benchmark CASIA-B. Extensive experiments show that EV-Gait can get nearly 96% recognition accuracy in the real-world settings, while on the CASIA-B benchmark it achieves comparable performance with state-of-the-art RGB-based gait recognition approaches.