CinePile: A Long Video Question Answering Dataset and Benchmark.
Ruchit RawalKhalid SaifullahRonen BasriDavid JacobsGowthami SomepalliTom GoldsteinPublished in: CoRR (2024)
Keyphrases
- question answering
- long video
- natural language
- information retrieval
- information extraction
- natural language processing
- natural language questions
- video clips
- qa clef
- question classification
- cross language
- passage retrieval
- question answering systems
- syntactic information
- video segments
- speech transcripts
- scene segmentation
- candidate answers
- image classification
- low level
- video sequences