Combining transcription-based and acoustic-based speaker identifications for broadcast news.
Elie el KhouryAntoine LaurentSylvain MeignierSimon PetitrenaudPublished in: ICASSP (2012)
Keyphrases
- broadcast news
- acoustic models
- automatic speech recognition
- speaker diarization
- speaker identification
- speech recognition
- speech recognition systems
- speech recognizer
- acoustic features
- speaker independent
- story segmentation
- mel frequency cepstral coefficients
- speech segments
- visual speech
- hidden markov models
- spoken document retrieval
- video search
- conversational speech
- video retrieval
- automatic transcription
- out of vocabulary
- spoken term detection
- spoken documents
- speech retrieval
- noisy environments
- image content
- image features
- search engine