Joint gaze-correction and beautification of DIBR-synthesized human face via dual sparse coding.
Xianming LiuGene CheungDeming ZhaiDebin ZhaoHiroshi SankohSei NaitoPublished in: ICIP (2014)
Keyphrases
- sparse coding
- human faces
- synthesized images
- sparse representation
- dictionary learning
- face recognition
- natural images
- lighting conditions
- face detection
- facial expressions
- unsupervised learning
- image classification
- eye tracking
- generative model
- image representation
- face images
- linear combination
- depth image based rendering
- discriminative power
- facial features
- feature space
- image patches
- visual words
- active appearance models
- high dimensional
- eye movements
- visual attention
- bag of words
- multiscale