Robust speech/non-speech detection using LDA applied to MFCC for continuous speech recognition.
Arnaud MartinGéraldine DamnatiLaurent MauuaryPublished in: INTERSPEECH (2001)
Keyphrases
- continuous speech recognition
- speech recognition
- speech signal
- noisy environments
- speaker identification
- detection algorithm
- vocal tract
- automatic speech recognition
- speaker recognition
- topic modeling
- audio visual
- partial occlusion
- linear prediction
- text to speech
- face recognition
- speech synthesis
- speaker diarization
- speech enhancement
- speech recognition systems
- detection method
- anomaly detection
- hidden markov models
- feature extraction
- information retrieval
- false alarms
- latent dirichlet allocation
- automatic detection
- spoken language
- discriminant analysis
- linear discriminant analysis
- topic models
- speech recognizer
- generative model
- voice activity detection