WREN: Low Light Image Enhancement Using Retinex theory-based Double U-Net-like Structures
Computer Vision and Pattern Recognition
Summary
The authors developed a new neural network called WREN to improve pictures taken in low light by separating the image into its color details and lighting using retinex theory. Their method uses two U-Net-like neural networks: one to split the image into reflectance and illumination parts, and another to enhance just the lighting part before putting them back together. This design helps avoid problems like over-brightening and works well across different lighting conditions. They trained the system to be robust to changes in brightness and showed that it performs better than previous methods on several tests.
Authors
Reina Kaneko, Junya Hara, Hiroshi Higashi, Yuichi Tanaka
Abstract
This paper proposes a neural network for low light image enhancement (LLIE) based on retinex theory to make LLIE robust for various dynamic range scenes. The retinex theory is an image formulation model inspired by a human color perception hypothesis, where a low light image is decomposed into intrinsic color context (i.e., reflectance map) and scene-dependent illumination (i.e., illumination map). Due to non-uniqueness of its decomposition, existing retinex-based LLIE methods often fail to achieve stable decomposition, which lead to over-enhancement. Typically, they are sensitive to the dynamic ranges that vary in different lighting conditions. To tackle this issue, we propose WREN: An LLIE neural network with double U-Net-like structures. WREN consists of two U-Net-like sub-networks. The first network has one encoder and two decoders that decompose an input image into the reflectance and illumination maps. The second network with a customized Transformer block between an encoder and a decoder only enhances the illumination map obtained from the first network: This completely follows the assumption of the retinex theory. Finally, the enhanced illumination map is recombined with the reflectance map. The network is trained end-to-end with a scale-invariant loss function, which gives robustness against the illumination scaling. Numerical results show that our method achieves the state-of-the-art performance across multiple datasets. Our code is available online.