MERLIN: Multimodal Embedding Refinement via LLM-based Iterative Navigation for Text-Video Retrieval-Rerank Pipeline.
Donghoon HanEunhwan ParkGisang LeeAdam LeeNojun KwakPublished in: CoRR (2024)
Keyphrases
- video retrieval
- video search
- concept based video retrieval
- video collections
- video segments
- visual content
- video database
- concept detection
- video indexing
- content based retrieval
- semantic gap
- video data
- information retrieval
- image and video retrieval
- multi modal
- key frames
- text mining
- content based video retrieval
- video content
- retrieval systems
- image search
- video clips
- text documents
- video shots
- broadcast news
- search engine
- semantic video retrieval
- interactive retrieval
- keywords
- video sequences
- image processing
- semantic concept detection
- semantic content