Revisiting loss-specific training of filter-based MRFs for image restoration

Abstract

It is now well known that Markov random fields (MRFs) are particularly effective for modeling image priors in low-level vision. Recent years have seen the emergence of two main approaches for learning the parameters in MRFs: (1) probabilistic learning using sampling-based algorithms and (2) loss-specific training based on MAP estimate. After investigating existing training approaches, it turns out that the performance of the loss-specific training has been significantly underestimated in existing work. In this paper, we revisit this approach and use techniques from bi-level optimization to solve it. We show that we can get a substantial gain in the final performance by solving the lower-level problem in the bi-level framework with high accuracy using our newly proposed algorithm. As a result, our trained model is on par with highly specialized image denoising algorithms and clearly outperforms probabilistically trained MRF models. Our findings suggest that for the loss-specific training scheme, solving the lower-level problem with higher accuracy is beneficial. Our trained model comes along with the additional advantage, that inference is extremely efficient. Our GPU-based implementation takes less than 1s to produce state-of-the-art performance.

10 pages, 2 figures, appear at 35th German Conference, GCPR 2013, Saarbrücken, Germany, September 3-6, 2013. Proceedings

Cited by in corpus (10)

• Trainable Nonlinear Reaction Diffusion: A Flexible Framework for Fast and Effective Image Restoration
• RENOIR - A Dataset for Real Low-Light Image Noise Reduction
• Insights into analysis operator learning: From patch-based sparse models to higher-order MRFs
• Cascades of Regression Tree Fields for Image Restoration
• The structure of optimal parameters for image restoration problems
• Bilevel methods for image reconstruction
• Discriminative Transfer Learning for General Image Restoration
• An abstract convergence framework with application to inertial inexact forward--backward methods
• Dyadic partition-based training schemes for TV/TGV denoising
• Learning optimal orders of the underlying Euclidean norm in total variation image denoising