End-to-End Speech Separation Using Orthogonal Representation in Complex and Real Time-Frequency Domain.
Kai WangHao HuangYing HuZhihua HuangSheng LiPublished in: Interspeech (2021)
Keyphrases
- end to end
- frequency domain
- real time
- spatial domain
- fourier transform
- text localization and recognition
- mathematical formalism
- cross correlation
- denoising
- feature extraction
- power spectrum
- congestion control
- multipath
- speech recognition
- affine invariance
- spectrum analysis
- filter bank
- fast fourier transform
- feature selection
- power spectra
- subband
- jpeg images
- image processing