Self-supervised Pretraining for Robust Personalized Voice Activity Detection in Adverse Conditions.
Holger Severin BovbjergJesper JensenJan ØstergaardZheng-Hua TanPublished in: CoRR (2023)
Keyphrases
- voice activity detection
- noisy environments
- sufficient conditions
- real world conditions
- information systems
- feature selection
- feature descriptors
- environmental conditions
- case study
- multiscale
- probabilistic model
- database
- computationally efficient
- speech recognition
- user modeling
- computer vision
- artificial intelligence
- neural network