Optimal Combination of Image Denoisers

Given a set of image denoisers, each having a different denoising capability, is there a provably optimal way of combining these denoisers to produce an overall better result? An answer to this question is fundamental to designing ensembles of weak estimators for complex scenes. In this paper, we present an optimal procedure leveraging deep neural networks and convex optimization. The proposed framework, called the Consensus Neural Network (CsNet), introduces three new concepts in image denoising: (1) A deep neural network to estimate the mean squared error (MSE) of denoised images without needing the ground truths; (2) A provably optimal procedure to combine the denoised outputs via convex optimization; (3) An image boosting procedure using a deep neural network to improve contrast and to recover lost details of the combined images. Experimental results show that CsNet can consistently improve denoising performance for both deterministic and neural network denoisers.