Robust Image Denoising of No-Flash Images Guided by Consistent Flash Images, Computer Graphics Lab. @ GIST

Robust Image Denoising of No-Flash Images Guided by Consistent Flash Images

Geunwoo Oh ¹ , Jonghee Back ¹ , Jae-Pil Heo ² , Bochang Moon ¹

GIST ¹ , Sungkyunkwan University ²

Thirty-Seventh AAAI Conference on Artificial Intelligence (AAAI) [Oral presentation]

Self-Supervised Post-Correction for Monte Carlo Denoising

Comparisons with the state-of-the-art techniques (DJF (Li et al. 2016) and CU-Net (Deng and Dragotti 2021)) that take a pair of flash/no-flash images as input. The no-flash images (a) are corrupted by Gaussian noise with σ = 75. The image regions in the flash images (b) contain specular highlights (in the top row) and hard shadows (in the bottom) that do not exist in the no-flash images ((a) and (f)). These inconsistent flash images make the existing methods ((c) and (d)) generate noticeable artifacts (residual noise and ghosting). On the other hand, our method produces much-reduced artifacts with higher numerical accuracy by robustly exploiting such inconsistent flash images.

Abstract

Images taken in low light conditions typically contain distracting noise, and eliminating such noise is a crucial computer vision problem. Additional photos captured with a camera flash can guide an image denoiser to preserve edges since the flash images often contain fine details with reduced noise. Nonetheless, a denoiser can be misled by inconsistent flash images, which have image structures (e.g., edges) that do not exist in no-flash images. Unfortunately, this disparity frequently occurs as the flash/no-flash pairs are taken in different light conditions. We propose a learning-based technique that robustly fuses the image pairs while considering their inconsistency. Our framework infers consistent flash image patches locally, which have similar image structures with the ground truth, and denoises no-flash images using the inferred ones via a combination model. We demonstrate that our technique can produce more robust results than state-of-the-art methods, given various flash/no-flash pairs with inconsistent image structures. The source code is available at https://github.com/CGLab-GIST/RIDFnF.