Wawenets: A No-Reference Convolutional Waveform-Based Approach to Estimating Narrowband and Wideband Speech Quality.
Andrew A. CatellierStephen D. VoranPublished in: ICASSP (2020)
Keyphrases
- speech quality
- speech signal
- speech recognition
- packet loss
- spectral analysis
- video quality
- image quality
- hidden markov models
- non stationary
- background noise
- quality estimation
- noisy environments
- voice over internet protocol
- adaptive filtering
- quality assessment
- sound source
- video content
- visual quality
- frequency domain
- quality metrics
- frequency band
- denoising
- multiscale
- image segmentation
- machine learning