IMFD IMPRESEE at TRECVID 2020: Description Generation by Visual-Syntactic Embedding.
Jesus Perez-MartinBenjamin BustosJorge PérezJuan Manuel BarriosPublished in: TRECVID (2020)
Keyphrases
- video search
- low level
- high level
- visual information
- video retrieval
- structural descriptions
- news video
- visual features
- natural language
- video data
- vector space
- semantic representation
- semantic indexing
- text generation
- video concept detection
- event detection
- representation scheme
- information retrieval
- generation process
- feature extraction