CODIS: Benchmarking Context-Dependent Visual Comprehension for Multimodal Large Language Models.
Fuwen LuoChi ChenZihao WanZhaolu KangQidong YanYingjie LiXiaolong WangSiyu WangZiyue WangXiaoyue MiPeng LiNing MaMaosong SunYang LiuPublished in: CoRR (2024)
Keyphrases
- context dependent
- language model
- relevance model
- low level
- language modeling
- probabilistic model
- speech recognition
- language modelling
- document retrieval
- n gram
- information retrieval
- visual information
- high level
- query expansion
- vector space model
- retrieval model
- statistical language models
- ad hoc information retrieval
- query terms
- context sensitive
- natural language
- test collection
- visual data
- audio visual
- multi modal
- language models for information retrieval
- smoothing methods
- multimedia
- pseudo relevance feedback
- visual features
- bayesian networks
- ad hoc retrieval
- natural language processing
- spoken term detection