A novel Reinforcement Learning (RL) method for generating context-based captions for videos. It would be fascinating to have it implemented in Julia