Evaluating Soccer Player: from Live Camera to Deep Reinforcement Learning.
Paul GarnierThéophane GregoirPublished in: CoRR (2021)
Keyphrases
- reinforcement learning
- soccer games
- real time
- field of view
- soccer game
- vision system
- camera calibration
- video camera
- camera images
- structure from motion
- live video
- robot soccer
- function approximation
- camera motion
- robocup soccer
- surveillance system
- optimal policy
- single camera
- hand held
- event detection
- optimal control
- camera parameters
- reinforcement learning algorithms
- dynamic programming
- multi agent
- focal length
- model free
- multiple cameras
- autocalibration
- action selection
- game play
- markov decision processes
- state space
- machine learning
- position and orientation
- ground plane
- imperfect information
- temporal difference learning
- video surveillance
- digital camera