MDMMT: Multidomain Multimodal Transformer for Video Retrieval.
Maksim DzabraevMaksim KalashnikovStepan KomkovAleksandr PetiushkoPublished in: CoRR (2021)
Keyphrases
- video retrieval
- visual content
- semantic gap
- video database
- content based retrieval
- video data
- video content
- video search
- retrieval systems
- concept detection
- image and video retrieval
- key frames
- video indexing
- multi modal
- concept based video retrieval
- video shots
- video collections
- interactive retrieval
- video clips
- video segments
- retrieval framework
- audio visual