Improving broadcast news transcription by lightly supervised discriminative training.
Ho Yin ChanPhilip C. WoodlandPublished in: ICASSP (1) (2004)
Keyphrases
- broadcast news
- discriminative training
- automatic speech recognition
- speech recognition
- acoustic models
- hidden markov models
- speaker identification
- speech signal
- spoken document retrieval
- speech recognizer
- learning algorithm
- probabilistic interpretation
- speaker diarization
- maximum likelihood
- unsupervised learning
- supervised learning
- video search
- neural network
- active learning
- three dimensional
- image processing
- machine learning