Deep Auto-Encoder Based on Supervised Learning for Damaged Face Reconstruction.
Ting RuiSai ZhangJunhua ZouYou ZhouJian TangPublished in: ICIMCS (2017)
Keyphrases
- supervised learning
- image reconstruction
- motion estimation
- bit rate
- statistical learning
- reconstructed image
- human faces
- high resolution
- unsupervised learning
- semi supervised learning
- reconstruction error
- multiple instance learning
- learning algorithm
- reconstruction method
- rate distortion
- image quality
- face images
- labeled data
- semi supervised
- reinforcement learning
- training data
- three dimensional
- decoding process
- discrete tomography
- compressive sensing
- video codec
- machine learning
- training set
- low complexity
- pairwise
- optical flow
- class labels