# DST-Net: A Dual-Stream Transformer with Illumination-Independent Feature Guidance and Multi-Scale Spatial Convolution for Low-Light Image Enhancement

arXiv: https://arxiv.org/abs/2603.16482 · HTML: https://mlsanity.com/paper/2603.16482/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 23 of its 55 references that are in the map.
- First posted on arXiv 17 Mar 2026.

## Builds on (most cited first, 23 of 23)

- [Squeeze-and-Excitation Networks](https://mlsanity.com/paper/1709.01507/) (2017)
- [CBAM: Convolutional Block Attention Module](https://mlsanity.com/paper/1807.06521/) (2018)
- [Deep Retinex Decomposition for Low-Light Enhancement](https://mlsanity.com/paper/1808.04560/) (2018)
- [Zero-Reference Deep Curve Estimation for Low-Light Image Enhancement](https://mlsanity.com/paper/2001.06826/) (2020)
- [EnlightenGAN: Deep Light Enhancement without Paired Supervision](https://mlsanity.com/paper/1906.06972/) (2019)
- [Rethinking Coarse-to-Fine Approach in Single Image Deblurring](https://mlsanity.com/paper/2108.05054/) (2021)
- [Learning Enriched Features for Real Image Restoration and Enhancement](https://mlsanity.com/paper/2003.06792/) (2020)
- [FFA-Net: Feature Fusion Attention Network for Single Image Dehazing](https://mlsanity.com/paper/1911.07559/) (2019)
- [Retinex-inspired Unrolling with Cooperative Prior Architecture Search for Low-light Image Enhancement](https://mlsanity.com/paper/2012.05609/) (2020)
- [Retinexformer: One-stage Retinex-based Transformer for Low-light Image Enhancement](https://mlsanity.com/paper/2303.06705/) (2023)
- [Attentional Feature Fusion](https://mlsanity.com/paper/2009.14082/) (2020)
- [R2RNet: Low-light Image Enhancement via Real-low to Real-normal Network](https://mlsanity.com/paper/2106.14501/) (2021)
- [You Only Need 90K Parameters to Adapt Light: A Light Weight Transformer for Image Enhancement and Exposure Correction](https://mlsanity.com/paper/2205.14871/) (2022)
- [HVI: A New Color Space for Low-light Image Enhancement](https://mlsanity.com/paper/2502.20272/) (2025)
- [Adaptive Unfolding Total Variation Network for Low-Light Image Enhancement](https://mlsanity.com/paper/2110.00984/) (2021)
- [LYT-NET: Lightweight YUV Transformer-based Network for Low-light Image Enhancement](https://mlsanity.com/paper/2401.15204/) (2024)
- [URWKV: Unified RWKV Model with Multi-state Perspective for Low-light Image Restoration](https://mlsanity.com/paper/2505.23068/) (2025)
- [Noise Self-Regression: A New Learning Paradigm to Enhance Low-Light Images Without Task-Related Data](https://mlsanity.com/paper/2211.04700/) (2022)
- [Low-Light Image Enhancement using Event-Based Illumination Estimation](https://mlsanity.com/paper/2504.09379/) (2025)
- [NTIRE 2025 Challenge on Video Quality Enhancement for Video Conferencing: Datasets, Methods and Results](https://mlsanity.com/paper/2505.18988/) (2025)
- [Geometric-Aware Low-Light Image and Video Enhancement via Depth Guidance](https://mlsanity.com/paper/2312.15855/) (2023)
- [UltraFast-LiNET: Light-weight multi-scale shift convolutional network for real-time low-light image enhancement](https://mlsanity.com/paper/2512.02965/) (2025)
- [Troublemaker Learning for Low-Light Image Enhancement](https://mlsanity.com/paper/2402.04584/) (2024)

## Abstract (arXiv)

Low-light image enhancement aims to restore the visibility of images captured by visual sensors in dim environments by addressing their inherent signal degradations, such as luminance attenuation and structural corruption. Although numerous algorithms attempt to improve image quality, existing methods often cause a severe loss of intrinsic signal priors. To overcome these challenges, we propose a Dual-Stream Transformer Network (DST-Net) based on illumination-agnostic signal prior guidance and multi-scale spatial convolutions. First, to address the loss of critical signal features under low-light conditions, we design a feature extraction module. This module integrates Difference of Gaussians (DoG), LAB color space transformations, and VGG-16 for texture extraction, utilizing decoupled illumination-agnostic features as signal priors to continuously guide the enhancement process. Second, we construct a dual-stream interaction architecture. By employing a cross-modal attention mechanism, the network leverages the extracted priors to dynamically rectify the deteriorated signal representation of the enhanced image, ultimately achieving iterative enhancement through differentiable curve estimation. Furthermore, to overcome the inability of existing methods to preserve fine structures and textures, we propose a Multi-Scale Spatial Fusion Block (MSFB) featuring pseudo-3D and 3D gradient operator convolutions. This module integrates explicit gradient operators to recover high-frequency edges while capturing inter-channel spatial correlations via multi-scale spatial convolutions. Extensive evaluations and ablation studies demonstrate that DST-Net achieves superior performance in subjective visual quality and objective metrics. Specifically, our method achieves a PSNR of 25.64 dB on the LOL dataset. Subsequent validation on the LSRW dataset further confirms its robust cross-scene generalization.