Low-light Raw Image Enhancement
Fig.1. Representative samples for low/normal-light RAW images. ’Challenging’ means that RAW images are captured under different lighting conditions.
Low-light Raw Image Enhancement
Compared with images under normal lighting conditions, the quality of low-light images captured under poor lighting conditions deteriorates severely due to unavoidable environmental or technical constraints, resulting in unpleasant visual perceptions including degradation of details, color distortion, and severe noise. To alleviate the problem of image quality degradation, low-light image enhancement has become an important topic in the field of underlying image processing to effectively improve visual quality and restore image details. In addition to low-light enhancements in the sRGB domain, RAW is also receiving more and more attention, which can keep a higher bit depths and linearity compared to sRGB images. This enhancement not only adapts to challenging lighting conditions but also can benefits various vision tasks.
We will use the low-light RAW image enhancement dataset proposed by Prof. Fu’s team in. They capture paired normal/low-light images dataset using a Canon EOS 5D Mark IV camera with different exposure ratios. The images are captured in a variety of indoor or outdoor scenes. It is noteworthy that all the scenes in their dataset are static to ensure that the content of the low-light image and its ground-truth are identical. We will host the competition using open source online platform, e.g. CodaLab. All submissions are evaluated by our script running on the server and we will double check the results of top-rank methods manually before releasing the final test-set rating.