Demonstration-Conditioned Reinforcement Learning for Few-Shot Imitation.
Christopher R. DanceJulien PerezThéo CachetPublished in: ICML (2021)
Keyphrases
- reinforcement learning
- function approximation
- model free
- video shots
- temporal difference
- state space
- video data
- reinforcement learning algorithms
- video sequences
- machine learning
- imitation learning
- video content
- key frames
- optimal control
- optimal policy
- detection algorithm
- visual features
- learning process
- multi agent
- supervised learning
- action selection
- partially observable
- reinforcement learning methods
- shot boundary detection
- neural network
- multimedia
- learning algorithm
- genetic algorithm
- policy search
- shot change detection