Fooling Deep Neural Networks
Anh Nguyen discusses the potential for fooling deep neural networks and the implications for security systems like facial recognition and self-driving cars. While larger scale networks can learn to recognize fooling images, smaller scale networks continue to be fooled even after multiple iterations of training. Future research will focus on investigating the fooling problem on different network architectures, including generative models.In this clip
From this podcast
Data Skeptic
Easily Fooling Deep Neural Networks
Related Questions