Multimodal — MOC
상위: VAULT-INDEX · 규칙: 설계 규약
비전-언어/임의모달 모델.
목차
- 대조 인코더: CLIP · SigLIP
- 브리징 VLM: Flamingo · BLIP-2 · LLaVA · CogVLM · MiniCPM-V
- 패밀리: Qwen-VL · InternVL · DeepSeek-VL2 · PaliGemma · Pixtral · Molmo
- 통합/any-to-any: Chameleon · Emu3 · Janus-Pro · GPT-4o
- 오디오: Whisper
키워드: #multimodal CLIP LLaVA Qwen-VL GPT-4o any-to-any
현재 노트 (25)
- [[20.AI/25.Multimodal/[CLIP]_형준|[CLIP]_형준]]
- blip-2
- chameleon
- clip
- cogvlm
- deepseek-vl2
- emu3
- flamingo
- gpt-4o
- internvl-3
- internvl
- janus-pro
- llava-onevision
- minicpm-v
- molmo
- paligemma-2
- phi-4-multimodal
- pixtral
- qwen2-vl
- qwen3-omni
- qwen3-vl
- show-o2
- siglip
- transfusion
- whisper