Pre‐trained low‐light image enhancement transformer

• Published in: IET image processing 2024-06, Vol.18 (8), p.1967-1984
• Main Authors: Zhang, Jingyao, Hao, Shijie, Rao, Yuan
• Format: Article
• Language: English
• Subjects: image enhancement; image processing
• ISSN: 1751-9659; 1751-9667
• DOI: 10.1049/ipr2.13076

Low‐light image enhancement is a longstanding challenge in low‐level vision, as images captured in low‐light conditions often suffer from significant aesthetic quality flaws. Recent methods based on deep neural networks have made impressive progress in this area. In contrast to mainstream convolutional neural network (CNN)‐based methods, an effective solution inspired by the transformer, which has shown impressive performance in various tasks, is proposed. This solution is centred around two key components. The first is an image synthesis pipeline, and the second is a powerful transformer‐based pre‐trained model, known as the low‐light image enhancement transformer (LIET). The image synthesis pipeline includes illumination simulation and realistic noise simulation, enabling the generation of more life‐like low‐light images to overcome the issue of data scarcity. LIET combines streamlined CNN‐based encoder‐decoders with a transformer body, efficiently extracting global and local contextual features at a relatively low computational cost. The extensive experiments show that this approach is highly competitive with current state‐of‐the‐art methods. The codes have been released and are available at LIET. An effective transformer‐based low‐light image enhancement solution called low‐light image enhancement transformer, pre‐trained on a large synthesized low/normal light image dataset, which achieves state‐of‐the‐art performance, is proposed. The model combines convolutional neural network and transformer architectures for robust feature extraction at low cost and improved generalization capability.