Adapt2Reward: Adapting Video-Language Models to Generalizable Robotic Rewards via Failure Prompts.
Yanting YangMinghao ChenQibo QiuJiahao WuWenxiao WangBinbin LinZiyu GuanXiaofei HePublished in: CoRR (2024)
Keyphrases
- language model
- reinforcement learning
- language modeling
- n gram
- bandit problems
- document retrieval
- probabilistic model
- speech recognition
- retrieval model
- video sequences
- language modelling
- video data
- reward function
- smoothing methods
- statistical language models
- information retrieval
- test collection
- markov decision processes
- multimedia
- context sensitive
- video content
- query expansion
- query terms
- expected reward
- vector space model
- video retrieval
- video frames
- ad hoc information retrieval
- document ranking
- language model for information retrieval
- cross lingual
- relevance model
- statistical language modeling
- language models for information retrieval
- spoken term detection