Kernel linear regression for low resolution face recognition under variable illumination.
Shih-Ming HuangJar-Ferr YangPublished in: ICASSP (2012)
Keyphrases
- linear regression
- low resolution
- face images
- face recognition
- kernel regression
- high resolution
- least squares
- image super resolution
- super resolution
- regression methods
- kernel density estimators
- human faces
- regression method
- low resolution face images
- low resolution images
- nonlinear regression
- high quality
- facial features
- principal component analysis
- lower resolution
- feature extraction
- facial expressions
- higher resolution
- discriminant analysis
- active appearance models
- super resolution reconstruction
- probe image
- linear discriminant analysis
- computer vision
- square loss
- support vector
- feature space
- face recognition algorithms
- image processing
- feature vectors
- kernel function
- high resolution images
- face detection
- sparse representation
- depth map
- face hallucination
- image sequences
- facial images
- motion estimation
- pixel intensities
- thermal images
- three dimensional