Bimodal log-linear regression for fusion of audio and visual features.
Ognjen RudovicStavros PetridisMaja PanticPublished in: ACM Multimedia (2013)
Keyphrases
- linear regression
- visual features
- visual information
- late fusion
- audio features
- visual data
- least squares
- acoustic features
- image classification
- visual content
- image retrieval
- low level features
- multiple features
- keywords
- regression methods
- image search
- nonlinear regression
- audio visual
- low level
- image collections
- image annotation
- key frames
- multimedia
- visual appearance
- content based video retrieval
- semantic concepts
- global features
- search engine
- visual similarity
- video shots
- probabilistic model
- high level
- computer vision