Fig. 3
From: Evaluating gait system vulnerabilities through PPO and GAN-generated adversarial attacks
Architecture of the proposed deep learning-based gait recognition model. This diagram delineates the multi-layered approach employed to extract and learn discriminative features from Gait Energy Images (GEIs), essential for precise individual identification. The illustrated CNN model is foundational for our experiments, serving as the pivotal component upon which our adversarial attack simulations are constructed and evaluated, underscoring its integral role in validating the efficacy of our proposed adversarial strategies.
