Relevance-based Margin for Contrastively-trained Video Retrieval Models.
Alex FalconSwathikiran SudhakaranGiuseppe SerraSergio EscaleraOswald LanzPublished in: ICMR (2022)
Keyphrases
- retrieval model
- test collection
- information retrieval
- relevance feedback
- language modeling
- document retrieval
- language model
- retrieval effectiveness
- training set
- information retrieval systems
- trec collections
- inference network model
- retrieval framework
- video data
- retrieval systems
- statistical language models
- vector space
- multimedia
- query topic
- vector space model
- video sequences
- video streams
- video content
- query expansion
- relevance judgments
- video frames
- retrieval functions
- average precision
- sentence retrieval
- user queries
- tf idf
- query terms
- video database
- video clips
- relevance model
- relevant documents
- video retrieval
- document collections
- search tasks
- language modeling framework
- key frames
- search engine
- probabilistic retrieval models
- text classification
- implicit feedback
- retrieved documents
- pseudo relevance feedback
- image retrieval