Improved Speech Separation with Time-and-Frequency Cross-Domain Feature Selection.
Tian LanYuxin QianYilan LyuRefuoe MokhosiWenxin TaiQiao LiuPublished in: Interspeech (2021)
Keyphrases
- cross domain
- feature selection
- text categorization
- web image annotation
- multi domain
- transfer learning
- knowledge transfer
- speech recognition
- domain adaptation
- sentiment classification
- cross domain learning
- multiple domains
- feature space
- feature set
- feature extraction
- machine learning
- text classification
- dimensionality reduction
- multi task
- support vector
- information systems
- margin maximization
- prior knowledge
- speech signal
- target domain
- neural network
- data sets