Meta-Reinforcement Learning by Tracking Task Non-stationarity.
Riccardo PoianiAndrea TirinzoniMarcello RestelliPublished in: IJCAI (2021)
Keyphrases
- reinforcement learning
- particle filter
- non stationary
- state space
- real time
- visual tracking
- real time tracking
- function approximation
- object tracking
- meta level
- learning problems
- motion tracking
- particle filtering
- video surveillance
- appearance model
- dynamic programming
- mean shift
- moving target
- video sequences
- model free
- multi agent
- reinforcement learning algorithms
- articulated objects
- neural network