Removing illumination artifacts from face images using the nuisance attribute projection.
Vitomir StrucBostjan VesnicerFrance MihelicNikola PavesicPublished in: ICASSP (2010)
Keyphrases
- face images
- illumination variations
- pose variations
- illumination compensation
- cmu pie
- face recognition
- face representation
- human faces
- illumination normalization
- facial appearance
- face recognition systems
- lighting conditions
- facial expressions
- principal component analysis
- feature vectors
- face databases
- automatic face
- low resolution
- robust face recognition
- illumination insensitive
- frontal pose
- feature extraction
- imaging conditions
- face verification
- probe image
- face recognition algorithms
- high resolution
- illumination conditions
- high quality
- input image
- feature points
- age estimation
- training set
- facial features
- single image
- face model
- face space
- face matching
- varying illumination
- human face recognition
- image processing
- correlation filters
- gender recognition
- recognition rate
- object tracking
- image set
- appearance variations
- local binary pattern
- feature space