Fetching the paper…

Look where you look! Saliency-guided Q-networks for generalization in visual Reinforcement Learning · Around