Double-Task Deep Q-Learning with Multiple Views.
Tingzhu BaiJianing YangJun ChenXian GuoXiangsheng HuangYuxing YaoPublished in: ICCV Workshops (2017)
Keyphrases
- multiple views
- multi view
- reinforcement learning
- viewpoint
- d objects
- single view
- cooperative
- multi agent
- function approximation
- state space
- multiple images
- multiple cameras
- point correspondences
- feature correspondences
- range images
- fundamental matrix
- learning algorithm
- dynamic scenes
- action selection
- consistency constraints
- camera images
- dynamic programming
- multiple viewpoints
- multiple range images
- ground plane
- optical flow