Intoxicated Speech Detection by Fusion of Speaker Normalized Hierarchical Features and GMM Supervectors.
Daniel BoneMatthew BlackMing LiAngeliki MetallinouSungbok LeeShrikanth S. NarayananPublished in: INTERSPEECH (2011)
Keyphrases
- speaker recognition
- mel frequency cepstral coefficients
- speaker identification
- gaussian mixture model
- feature vectors
- speech recognition
- speech signal
- false positives
- speaker verification
- feature space
- speaker diarization
- acoustic features
- feature extraction
- voice activity detection
- automatic speech recognition
- audio visual
- linear predictive
- multiple features
- noisy environments
- feature set
- em algorithm
- audio features
- spectral features
- low level
- classification accuracy
- image sequences
- lexical features
- image retrieval
- hidden markov models
- speech synthesis
- text to speech
- information fusion
- image features
- detection algorithm
- data fusion