ROLL: Visual Self-Supervised Reinforcement Learning with Object Reasoning.
Yufei WangGautham Narayan NarasimhanXingyu LinBrian OkornDavid HeldPublished in: CoRR (2020)
Keyphrases
- reinforcement learning
- visual objects
- visual appearance
- spatial relations
- visual properties
- visual features
- d objects
- object model
- function approximation
- visual information
- knowledge representation
- state space
- learning algorithm
- visual input
- knowledge base
- low level
- markov decision processes
- automated reasoning
- temporal difference
- neural network
- keypoints
- image sequences
- visual scene
- real world objects
- multi agent
- multiple objects
- multi agent systems
- object segmentation
- object classes
- object recognition
- moving objects
- data objects