You’ve presumably noticed the multiple biometrics on display in my “landscape” reel. Fingers, a face, irises, a voice, veins, and even DNA.
But you may have missed one.
Gait.
I still remember the time that I ran into a former coworker from my pre-biometric years, and she said she recognized me by the way I walk.
But it’s no surprise that gait recognition is susceptible to spoofing.
“[A]dversarial Gait Recognition has arisen as a major challenge in video surveillance systems, as deep learning-based gait recognition algorithms become more sensitive to adversarial attacks.”
So Zeeshan Ali and others are working on IMPROVING gait-based adversarial attacks…the better to counter them.
“Our technique includes two major components: AdvHelper, a surrogate model that simulates the target, and PerturbGen, a latent-space perturbation generator implemented in an encoder-decoder framework. This design guarantees that adversarial samples are both effective and perceptually realistic by utilizing reconstruction and perceptual losses. Experimental results on the benchmark CASIA-gait dataset show that the proposed method achieves a high attack success rate of 94.33%.”
Now we need to better detect these adversarial attacks.
Bredemarket 