Speaker discrimination using formant trajectories from casework recordings: Can LDA do it?
Radek SkarnitzlJitka VankováPublished in: ICPhS (2015)
Keyphrases
- speech recognition
- audio visual
- automatic speech recognition
- linear discriminant analysis
- latent dirichlet allocation
- acoustic features
- vocal tract
- speech signal
- face recognition
- topic models
- topic modeling
- speaker adaptation
- spontaneous speech
- speaker verification
- generative model
- dimensionality reduction
- discriminant analysis
- principal component analysis
- moving objects
- speech synthesis
- dimension reduction
- feature selection
- trajectory data
- linear discriminate analysis
- topic discovery
- hidden markov models
- noisy environments
- language model
- multi modal
- discriminating power
- feature extraction
- speaker recognition
- speaker identification
- correlation filters
- bayesian networks
- pattern recognition
- co occurrence
- speaker independent
- feature space
- broadcast news
- speech recognizer
- motion patterns
- latent dirichlet
- latent semantic analysis