SPHINX: The Joint Mixing of Weights, Tasks, and Visual Embeddings for Multi-modal Large Language Models.
Ziyi LinChris LiuRenrui ZhangPeng GaoLongtian QiuHan XiaoHan QiuChen LinWenqi ShaoKeqin ChenJiaming HanSiyuan HuangYichi ZhangXuming HeHongsheng LiYu QiaoPublished in: CoRR (2023)
Keyphrases
- multi modal
- language model
- cross modal
- language modeling
- probabilistic model
- video search
- n gram
- document retrieval
- language modelling
- information retrieval
- speech recognition
- retrieval model
- statistical language models
- visual information
- single modality
- test collection
- semantic concepts
- multi modality
- context sensitive
- vector space model
- audio visual
- query expansion
- smoothing methods
- relevance model
- visual features
- high dimensional
- multiple modalities
- translation model
- vector space
- low level
- document ranking
- language models for information retrieval