IAML: Illumination-Aware Mirror Loss for Progressive Learning in Low-Light Image Enhancement Auto-encoders
This addresses image quality improvement in low-light conditions for applications like photography or surveillance, but it is incremental as it builds on existing auto-encoder and distillation methods.
The paper tackles low-light image enhancement by proposing a teacher-student auto-encoder with a progressive learning approach and a new Illumination-Aware Mirror Loss (IAML), achieving state-of-the-art performance on SSIM, PSNR, and LPIPS metrics across three datasets.
This letter presents a novel training approach and loss function for learning low-light image enhancement auto-encoders. Our approach revolves around the use of a teacher-student auto-encoder setup coupled to a progressive learning approach where multi-scale information from clean image decoder feature maps is distilled into each layer of the student decoder in a mirrored fashion using a newly-proposed loss function termed Illumination-Aware Mirror Loss (IAML). IAML helps aligning the feature maps within the student decoder network with clean feature maps originating from the teacher side while taking into account the effect of lighting variations within the input images. Extensive benchmarking of our proposed approach on three popular low-light image enhancement datasets demonstrate that our model achieves state-of-the-art performance in terms of average SSIM, PSNR and LPIPS reconstruction accuracy metrics. Finally, ablation studies are performed to clearly demonstrate the effect of IAML on the image reconstruction accuracy.