A regularization framework for stereo matching using IGMRF prior and sparseness learned from autoencoder.
Sonam NaharManjunath V. JoshiPublished in: ICIP (2016)
Keyphrases
- gaussian markov random field
- regularization framework
- autoregressive
- prior model
- super resolution
- edge preserving
- semi supervised learning
- semi supervised
- prior knowledge
- metric learning
- weighted graph
- gray scale
- image restoration
- least squares
- energy function
- unlabeled data
- labeled and unlabeled data
- high quality
- low resolution
- bilateral filter
- high resolution
- low spatial resolution
- multiscale