Deep Guided Exposure Correction with Knowledge Distillation
Lists
Only you see this
Abstract
Images captured with unreasonable exposures will greatly reduce visual quality. Exposure problems can be categorized as follows: (i) over-exposure, i.e., bright and losing image regions caused by too-long exposure; (ii) under-exposure, i.e., dark and drowned-in noises caused by too-short exposure. Most prior works only handle over- or under-exposure problems on sRGB domain and ignore prior knowledge of channel information. In this paper, we propose <b>D</b>eep <b>G</b>uided network for exposure correction on RAW domain with <b>K</b>nowledge <b>D</b>istillation (denoted as DGKD), solving two problems together. Firstly, according to color sensitivity, we employ blue/red channel and green channel as guidance information for over- and under-exposure correction, respectively. Secondly, to handle two varying problems in a unified network, we first train the over- and under-exposure correction networks individually and then distill knowledge into one deep guided network. The experimental results show that the proposed method outperforms the state-of-the-art methods under both quantitative metrics and visual quality. Specifically, the proposed method attained a peak signal-to-noise ratio of 24.653 dB and a structural similarity index of 0.8182 on the collected RAW image exposure correction dataset.
BibTeX
@inproceedings{liu2025deep,
author = {Sai Liu and Tao Zhang},
title = {{Deep Guided Exposure Correction with Knowledge Distillation}},
booktitle = {Sensors},
year = {2025},
doi = {10.3390/s25247606},
}Loading related content...