Disentangling Controllable Object Through Video Prediction Improves Visual Reinforcement Learning

Keyphrases

• Video Prediction 100%
• Visual Reinforcement Learning 100%
• Popular 50%
• Video Games 50%
• Performance Efficiency 50%
• Sampling Efficiency 50%
• Vision-based 50%
• Visual Field 50%
• Robotic Arm 50%
• Reinforcement Learning 50%
• Game Performance 50%
• Scoring Game 50%
• Agent-based Control 50%
• Movable Objects 50%
• Observation Channel 50%
• Reinforcement Learning Problems 50%
• Atari Games 50%
• Disentangled Representation 50%
• Double DQN Algorithm 50%

Computer Science

• Reinforcement Learning 100%
• videogame 33%
• Learning Problem 33%

Engineering

• Reinforcement Learning 100%
• Robotic Arm 33%