Interpretable Learning for Self-Driving Cars by Visualizing Causal Attention

Jinkyu Kim, John Canny

Deep neural perception and control networks are likely to be a key component of self-driving vehicles. These models need to be explainable - they should provide easy-to-interpret r…