Two-stage feature compensation of clean and telephone speech signals employing bidirectional neural network.
Iman EsmailiMansoor ValiJahanshah KabudianPublished in: ISSPA (2010)
Keyphrases
- speech signal
- neural network
- speech recognition
- noisy speech
- vocal tract
- automatic speech recognition
- speaker identification
- adaptive filtering
- non stationary
- artificial neural networks
- sound signals
- hidden markov models
- automatic speech recognition systems
- image processing
- additive noise
- spectral analysis
- speech enhancement
- background noise
- ultrasonic flow measurement
- sound source
- linear prediction
- fundamental frequency
- feature vectors
- feature selection
- genetic algorithm
- formant frequencies
- information retrieval