X-TaSNet: Robust and Accurate Time-Domain Speaker Extraction Network.
Zining ZhangBingsheng HeZhenjie ZhangPublished in: INTERSPEECH (2020)
Keyphrases
- computationally efficient
- highly accurate
- network model
- network structure
- peer to peer
- network traffic
- high accuracy
- learning algorithm
- frequency domain
- speech recognition
- completely automatic
- neural network
- automatic extraction
- partial occlusion
- maximum likelihood
- information extraction
- high quality
- social networks
- genetic algorithm