Missing-Feature Reconstruction by Leveraging Temporal Spectral Correlation for Robust Speech Recognition in Background Noise Conditions.
Wooil KimJohn H. L. HansenPublished in: IEEE Trans. Speech Audio Process. (2010)
Keyphrases
- speech recognition
- background noise
- noisy environments
- speech signal
- cepstral coefficients
- spectral subtraction
- speaker identification
- spectral analysis
- speech processing
- hidden markov models
- language model
- automatic speech recognition
- linear prediction
- noisy speech
- speech synthesis
- pattern recognition
- speech enhancement
- speech recognizer
- speech recognition technology
- information retrieval
- speaker adaptation
- speech recognition systems
- vocal tract
- noise reduction
- adaptive filtering
- low contrast
- word recognition
- high resolution
- image enhancement
- image analysis
- speaker dependent