Adding controlled amount of noise to improve recognition of compressed and spectrally distorted speech.
Jan NouzaPetr CervaJan SilovskýPublished in: ICASSP (2013)
Keyphrases
- noisy environments
- recognition rate
- recognition engine
- data structure
- speech recognition
- pattern recognition
- feature extraction
- action recognition
- noisy speech
- noisy data
- automatic recognition
- automatic speech recognition systems
- anti noise
- speech enhancement
- background noise
- random noise
- speech signal
- recognition accuracy
- image quality