Histogram equalization of real and imaginary modulation spectra for noise-robust speech recognition.
Hsin-Ju HsiehBerlin ChenJeih-weih HungPublished in: INTERSPEECH (2013)
Keyphrases
- speech recognition
- noisy environments
- histogram equalization
- image enhancement
- hidden markov models
- speech processing
- speech signal
- automatic speech recognition
- speech enhancement
- language model
- gray scale
- speech synthesis
- noisy speech
- speech recognizer
- background noise
- contrast enhancement
- noise reduction
- speaker identification
- image contrast
- gray level
- pattern recognition
- image contrast enhancement
- speech recognition systems
- speaker independent
- speech recognition technology
- word recognition
- signal to noise ratio
- similarity measure
- image processing
- information retrieval
- neural network