Lightweight Low-Light Image Enhancement via Distribution-Normalizing Preprocessing and Depthwise U-Net
This work addresses the need for efficient low-light image enhancement models with reduced computational cost, though the improvement is incremental over existing methods.
The authors propose a lightweight two-stage framework for low-light image enhancement that combines distribution-normalizing preprocessing with a depthwise-separable U-Net, achieving competitive perceptual quality with significantly fewer parameters and securing 4th place in the CVPR 2026 NTIRE Efficient Low-Light Image Enhancement Challenge.
We present a lightweight two-stage framework for low-light image enhancement (LLIE) that achieves competitive perceptual quality with significantly fewer parameters than existing methods. Our approach combines frozen algorithm-based preprocessing with a compact U-Net built entirely from depthwise-separable convolutions. The preprocessing normalizes the input distribution by providing complementary brightness-corrected views, enabling the trainable network to focus on residual color correction. Our method achieved 4th place in the CVPR 2026 NTIRE Efficient Low-Light Image Enhancement Challenge. We further provide extended benchmarks and ablations to demonstrate the general effectiveness of our methods.