Subspace-Based Speech Correlation Vector Estimation for Single-Microphone Multi-Frame MVDR Filtering.
Dörte FischerSimon DocloPublished in: ICASSP (2020)
Keyphrases
- multi frame
- linear subspace
- filtering algorithm
- automatic speech recognition
- image sequences
- motion estimation
- speech recognition
- low dimensional
- point correspondences
- speaker diarization
- computationally efficient
- principal component analysis
- feature vectors
- temporal coherence
- gray scale
- range images
- super resolution
- motion analysis
- dimensionality reduction
- high quality
- single frame
- feature extraction