Video Prediction Models as Rewards for Reinforcement Learning.
Alejandro EscontrelaAdemi AdenijiWilson YanAjay JainXue Bin PengKen GoldbergYoungwoon LeeDanijar HafnerPieter AbbeelPublished in: NeurIPS (2023)
Keyphrases
- reinforcement learning
- video data
- models built
- real time
- probabilistic model
- markov decision processes
- prediction model
- multimedia
- state space
- complex systems
- video streams
- video sequences
- dynamic programming
- reinforcement learning algorithms
- statistical models
- function approximation
- neural network
- video content
- video frames
- learning algorithm
- multi agent