# Interpretable Unsupervised Joint Denoising and Enhancement for Real-World low-light Scenarios

arXiv: https://arxiv.org/abs/2503.14535 · HTML: https://mlsanity.com/paper/2503.14535/

## Key facts

- Cited by 11 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 17 Mar 2025.

## Built on by (most cited first, 11 of 11)

- [Does Your Reasoning Model Implicitly Know When to Stop Thinking?](https://mlsanity.com/paper/2602.08354/) (2026)
- [Adaptive Batch-Wise Sample Scheduling for Direct Preference Optimization](https://mlsanity.com/paper/2506.17252/) (2025)
- [Heterogeneous Agent Collaborative Reinforcement Learning](https://mlsanity.com/paper/2603.02604/) (2026)
- [Real-Time Aligned Reward Model beyond Semantics](https://mlsanity.com/paper/2601.22664/) (2026)
- [IGDNet: Zero-Shot Robust Underexposed Image Enhancement via Illumination-Guided and Denoising](https://mlsanity.com/paper/2507.02445/) (2025)
- [Unsupervised Ultra-High-Resolution UAV Low-Light Image Enhancement: A Benchmark, Metric and Framework](https://mlsanity.com/paper/2509.01373/) (2025)
- [Anchor then Polish for Low-light Enhancement](https://mlsanity.com/paper/2603.15472/) (2026)
- [Internally Referenced Low-Light Enhancement](https://mlsanity.com/paper/2605.28605/) (2026)
- [FRBNet: Revisiting Low-Light Vision through Frequency-Domain Radial Basis Network](https://mlsanity.com/paper/2510.23444/) (2025)
- [EchoStyle: Unlocking High-Fidelity Video Stylization with Reverse Data Synthesis](https://mlsanity.com/paper/2606.25465/) (2026)
- [Learning Structural Illumination for Unsupervised Low-light Enhancement](https://mlsanity.com/paper/2608.08153/) (2026)

## Abstract (arXiv)

Real-world low-light images often suffer from complex degradations such as local overexposure, low brightness, noise, and uneven illumination. Supervised methods tend to overfit to specific scenarios, while unsupervised methods, though better at generalization, struggle to model these degradations due to the lack of reference images. To address this issue, we propose an interpretable, zero-reference joint denoising and low-light enhancement framework tailored for real-world scenarios. Our method derives a training strategy based on paired sub-images with varying illumination and noise levels, grounded in physical imaging principles and retinex theory. Additionally, we leverage the Discrete Cosine Transform (DCT) to perform frequency domain decomposition in the sRGB space, and introduce an implicit-guided hybrid representation strategy that effectively separates intricate compounded degradations. In the backbone network design, we develop retinal decomposition network guided by implicit degradation representation mechanisms. Extensive experiments demonstrate the superiority of our method. Code will be available at https://github.com/huaqlili/unsupervised-light-enhance-ICLR2025.