HiRes-LLaVA: Restoring Fragmentation Input in High-Resolution Large Vision-Language Models.
Runhui HuangXinpeng DingChunwei WangJianhua HanYulong LiuHengshuang ZhaoHang XuLu HouWei ZhangXiaodan LiangPublished in: CoRR (2024)
Keyphrases
- language model
- high resolution
- language modeling
- n gram
- document retrieval
- speech recognition
- language modelling
- image processing
- probabilistic model
- information retrieval
- statistical language models
- retrieval model
- computer vision
- query expansion
- language models for information retrieval
- query terms
- context sensitive
- smoothing methods
- test collection
- ad hoc information retrieval
- pseudo relevance feedback
- relevance model
- passage retrieval
- document ranking
- retrieval effectiveness
- term dependencies
- okapi bm
- hidden markov models