Superresolution-based stereo signal separation via supervised nonnegative matrix factorization.
Daichi KitamuraHiroshi SaruwatariYusuke IwaoKiyohiro ShikanoKazunobu KondoYu TakahashiPublished in: DSP (2013)
Keyphrases
- nonnegative matrix factorization
- super resolution
- low resolution
- depth map
- high resolution
- matrix factorization
- image reconstruction
- resolution enhancement
- low resolution images
- data representation
- negative matrix factorization
- high resolution images
- image restoration
- semi supervised
- signal processing
- stereo matching
- high quality
- least squares
- principal component analysis
- single image
- stereo images
- computer vision
- supervised learning
- frequency domain
- discontinuity preserving
- unsupervised learning
- spectral clustering
- motion estimation
- three dimensional
- feature selection
- face recognition
- collaborative filtering
- sparse representation
- original data
- objective function
- document clustering
- low dimensional
- document collections
- data sets