🤖 AI Summary
Existing methods typically address under-exposed or over-exposed images separately, lacking a unified correction framework, and suffer from three key limitations: weak local adaptivity, ill-posed regional exposure estimation, and inconsistent multi-exposure outputs for identical content. This paper proposes the first multi-dimensional regression framework for wide-range exposure correction. It constructs supervision signals via multi-exposure image generation, introduces a region-level exposure value regression module, and employs an attention-enhanced neural process (ANP) for pixel-wise adaptive refinement. Crucially, we pioneer a region-aware exposure prediction guidance mechanism to ensure both visual consistency and physical interpretability. On standard benchmarks, our method achieves state-of-the-art PSNR performance, with single-image inference taking only 0.05 seconds on an RTX 3090 GPU. Qualitatively, results are natural, robust, and generalizable across diverse exposure conditions.
📝 Abstract
Photo exposure correction is widely investigated, but fewer studies focus on correcting under- and over-exposed images simultaneously. Three issues remain open to handle and correct both under- and over-exposed images in a unified way. First, a locally-adaptive exposure adjustment may be more flexible instead of learning a global mapping. Second, it is an ill-posed problem to determine the suitable exposure values locally. Third, photos with the same content but different exposures may not reach consistent adjustment results. To this end, we proposed a novel exposure correction network, ExReg, to address the challenges by formulating exposure correction as a multi-dimensional regression process. Given an input image, a compact multi-exposure generation network is introduced to generate images with different exposure conditions for multi-dimensional regression and exposure correction in the next stage. An auxiliary module is designed to predict the region-wise exposure values, guiding the proposed Encoder-Decoder ANP (Attentive Neural Processes) to regress the final corrected image. The experimental results show that ExReg can generate well-exposed results and outperform the SOTA method in PSNR for extensive exposure problems. Furthermore, the processing speed, with 0.05 seconds per image on an RTX 3090, is efficient. When tested on the same image under various exposure levels, ExReg also yields results that are visually consistent and physically accurate.
