Contour Analysis Based on Empirical Mode Decomposition for DNN Acoustic Modeling in Mandarin Speech Recognition.
Xiaoyun WangXugang LuHisashi KawaiSeiichi YamamotoPublished in: INTERSPEECH (2016)
Keyphrases
- speech recognition
- speech recognition systems
- speaker independent
- hidden markov models
- speech synthesis
- speech recognizers
- speech recognizer
- empirical mode decomposition
- language model
- pattern recognition
- automatic speech recognition
- speech signal
- non stationary
- image analysis
- speech recognition technology
- noisy environments
- neural network
- speaker identification
- image quality
- signal processing
- image segmentation