Comparison of low- and high-level visual features for audio-visual continuous automatic speech recognition.
Petar S. AleksicAggelos K. KatsaggelosPublished in: ICASSP (5) (2004)
Keyphrases
- audio visual
- visual features
- visual information
- automatic speech recognition
- low and high level
- low level
- visual data
- acoustic features
- speech recognition
- audio features
- image classification
- visual content
- speaker verification
- multi modal
- speech signal
- image search
- low level features
- hidden markov models
- image annotation
- image retrieval
- noisy environments
- broadcast news
- keywords
- computer vision
- high level
- image collections
- key frames
- emotion recognition
- pattern recognition
- feature set
- object recognition
- multimedia