Tree-Structured Policy Based Progressive Reinforcement Learning for Temporally Language Grounding in Video.
Jie WuGuanbin LiSi LiuLiang LinPublished in: AAAI (2020)
Keyphrases
- reinforcement learning
- optimal policy
- policy search
- temporal information
- video data
- markov decision process
- action selection
- video sequences
- xml files
- reinforcement learning problems
- function approximation
- programming language
- markov decision processes
- partially observable
- tree structured data
- reinforcement learning algorithms
- tree structure
- video content
- language learning
- policy evaluation
- policy iteration
- action space
- function approximators
- control policies
- multimedia
- space time
- partially observable environments
- video streams
- video retrieval
- control policy
- natural language
- model free
- pattern languages
- markov decision problems
- partially observable domains
- multi agent
- spatio temporal
- rooted trees
- key frames
- state space
- video frames
- partially observable markov decision processes
- tree structures
- real time
- state action
- state and action spaces
- video clips
- policy gradient
- partially observable markov decision process
- temporal order
- optimal control
- reinforcement learning methods
- long run
- video analysis
- multimedia data
- decision problems
- structured data
- query language
- dynamic programming
- xml documents
- machine learning