Speech enhancement in presence of diffuse background noise: Why using blind signal extraction?
Jani EvenHiroshi SaruwatariKiyohiro ShikanoTomoya TakataniPublished in: ICASSP (2010)
Keyphrases
- background noise
- speech enhancement
- noisy environments
- noisy speech
- single channel
- speech signal
- noise reduction
- low contrast
- speech recognition
- signal processing
- signal to noise ratio
- multi channel
- additive noise
- linear prediction
- hidden markov models
- spectral subtraction
- neural network
- automatic speech recognition
- frequency domain
- non stationary
- pattern recognition