Joint estimation of late reverberant and speech power spectral densities in noisy environments using frobenius norm.
Ofer SchwartzSharon GannotEmanuël A. P. HabetsPublished in: EUSIPCO (2016)
Keyphrases
- joint estimation
- noisy environments
- speech signal
- power spectral
- frobenius norm
- speech recognition
- kullback leibler divergence
- speech enhancement
- power spectrum
- low rank
- reconstruction error
- frequency domain
- euclidean distance
- automatic speech recognition
- probability density function
- distance measure
- vocal tract
- speaker identification
- negative matrix factorization
- linear prediction
- non stationary
- voice activity detection
- hidden markov models
- matrix completion
- natural images
- probability distribution
- speech synthesis
- linear combination
- probability density
- matrix factorization
- singular value decomposition
- pattern recognition
- object recognition