Histogram-based spectral equalization for HMM-based speech synthesis using mel-LSP.
Yamato OhtaniMasatsune TamuraMasahiro MoritaTakehiko KagoshimaMasami AkaminePublished in: INTERSPEECH (2012)
Keyphrases
- speech synthesis
- speech recognition
- cepstral coefficients
- hidden markov models
- prosodic features
- vocal tract
- text to speech
- speech signal
- noisy environments
- language model
- speech corpus
- pattern recognition
- mean shift
- computer vision
- speaker identification
- linear prediction
- information retrieval
- spectral analysis
- automatic speech recognition
- contrast enhancement
- multipath
- particle filter