The Visual Centrifuge: Model-Free Layered Video Representations.
Jean-Baptiste AlayracJoão CarreiraAndrew ZissermanPublished in: CVPR (2019)
Keyphrases
- model free
- mid level
- reinforcement learning
- visual cues
- function approximation
- reinforcement learning algorithms
- visual analysis
- visual data
- video data
- visual information
- video sequences
- low level
- video content
- temporal difference
- video search
- multimedia
- news video
- video frames
- visual features
- policy iteration
- policy evaluation
- learning algorithm