Unsupervised Speaker Identification in TV Broadcast Based on Written Names.
Johann PoignantLaurent BesacierGeorges QuénotPublished in: IEEE ACM Trans. Audio Speech Lang. Process. (2015)
Keyphrases
- speaker identification
- tv broadcast
- feature transformation
- gaussian mixture model
- speech recognition
- feature extraction
- broadcast news
- speech signal
- unsupervised learning
- noisy environments
- semi supervised
- supervised learning
- named entities
- hidden markov models
- image processing
- keywords
- em algorithm
- image data
- multimedia