Robust multichannel TDOA estimation for speaker localization using the impulsive characteristics of speech spectrum.
Hongsen HeJingdong ChenJacob BenestyYingyue ZhouTao YangPublished in: ICASSP (2017)
Keyphrases
- localization method
- speech recognition
- speaker recognition
- audio visual
- robust estimation
- speaker verification
- speaker identification
- automatic speech recognition
- robust regression
- noisy environments
- speaker diarization
- speech signal
- object localization
- estimation error
- noise reduction
- gaussian mixture model
- text to speech
- prosodic features
- computationally efficient
- automatic speech recognition systems
- linear prediction
- multi channel
- parameter estimation