Combining Log-Spectral Domain Compensation with MVA Feature Post-Processing for Robust Speech Recognition.
Jianjun LeiJian WangJun GuoGang LiuHaifeng ShenPublished in: IIH-MSP (2006)
Keyphrases
- post processing
- speech recognition
- spectral domain
- noisy environments
- preprocessing
- hidden markov models
- language model
- speech signal
- pattern recognition
- speech recognition systems
- feature vectors
- frequency domain
- image compression
- human experts
- spatial domain
- probabilistic model
- optical coherence tomography
- neural network