Adversarial Robustness in AI
The conversation delves into the challenges posed by adversarial examples, particularly in the context of autonomous vehicles. Dragomir highlights the importance of using multiple sensors to enhance robustness and discusses innovative methods for unsupervised domain adaptation that improve performance across varying conditions. He emphasizes the balance between maintaining performance in standard scenarios while enhancing resilience against adversarial attacks.In this clip
From this podcast
Gradient Dissent - A Machine Learning Podcast
Drago Anguelov — Robustness, Safety, and Scalability at Waymo
Related Questions