||Deep reinforcement learning (DRL) algorithms provide a framework for automatic robot perception and control. In recent years, methods based on DRL have achieved great performance in different control tasks like grasping and locomotion. However, the question of how to make robots learn appropriate social behaviors under modern frameworks remains underexplored, partly due to the lack of cross-disciplinary synergies in human-robot interaction studies. As a consequence, the interaction scenarios studied in previous research have been limited to simplified cases and the algorithms studied to relatively simple ones. In this seminar, I would like to introduce my journey to solve this problem, from the initial human-robot interaction experiments to a recent deep reinforcement learning framework we considered.