# UltraFast-LiNET: Light-weight multi-scale shift convolutional network for real-time low-light image enhancement

arXiv: https://arxiv.org/abs/2512.02965 · HTML: https://mlsanity.com/paper/2512.02965/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Builds on 40 of its 69 references that are in the map.
- First posted on arXiv 2 Dec 2025.

## Builds on (most cited first, 30 of 40)

- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [MobileNetV2: Inverted Residuals and Linear Bottlenecks](https://mlsanity.com/paper/1801.04381/) (2018)
- [Aggregated Residual Transformations for Deep Neural Networks](https://mlsanity.com/paper/1611.05431/) (2016)
- [Xception: Deep Learning with Depthwise Separable Convolutions](https://mlsanity.com/paper/1610.02357/) (2016)
- [ShuffleNet: An Extremely Efficient Convolutional Neural Network for Mobile Devices](https://mlsanity.com/paper/1707.01083/) (2017)
- [Multi-Scale Context Aggregation by Dilated Convolutions](https://mlsanity.com/paper/1511.07122/) (2015)
- [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)
- [GhostNet: More Features from Cheap Operations](https://mlsanity.com/paper/1911.11907/) (2019)
- [Dynamic Convolution: Attention over Convolution Kernels](https://mlsanity.com/paper/1912.03458/) (2019)
- [Retinex-inspired Unrolling with Cooperative Prior Architecture Search for Low-light Image Enhancement](https://mlsanity.com/paper/2012.05609/) (2020)
- [LLNet: A Deep Autoencoder Approach to Natural Low-light Image Enhancement](https://mlsanity.com/paper/1511.03995/) (2015)
- [Learning to Enhance Low-Light Image via Zero-Reference Deep Curve Estimation](https://mlsanity.com/paper/2103.00860/) (2021)
- [Low-Light Image and Video Enhancement Using Deep Learning: A Survey](https://mlsanity.com/paper/2104.10729/) (2021)
- [Run, Don't Walk: Chasing Higher FLOPS for Faster Neural Networks](https://mlsanity.com/paper/2303.03667/) (2023)
- [CondenseNet: An Efficient DenseNet using Learned Group Convolutions](https://mlsanity.com/paper/1711.09224/) (2017)
- [R2RNet: Low-light Image Enhancement via Real-low to Real-normal Network](https://mlsanity.com/paper/2106.14501/) (2021)
- [MobileNetV4 -- Universal Models for the Mobile Ecosystem](https://mlsanity.com/paper/2404.10518/) (2024)
- [Shift: A Zero FLOP, Zero Parameter Alternative to Spatial Convolutions](https://mlsanity.com/paper/1711.08141/) (2017)
- [OneRestore: A Universal Restoration Framework for Composite Degradation](https://mlsanity.com/paper/2407.04621/) (2024)
- [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)
- [PCANet: A Simple Deep Learning Baseline for Image Classification?](https://mlsanity.com/paper/1404.3606/) (2014)
- [Retinexmamba: Retinex-based Mamba for Low-light Image Enhancement](https://mlsanity.com/paper/2405.03349/) (2024)
- [Adaptive Unfolding Total Variation Network for Low-Light Image Enhancement](https://mlsanity.com/paper/2110.00984/) (2021)
- [Wave-Mamba: Wavelet State Space Model for Ultra-High-Definition Low-Light Image Enhancement](https://mlsanity.com/paper/2408.01276/) (2024)
- [Loss Functions for Neural Networks for Image Processing](https://mlsanity.com/paper/1511.08861/) (2015)
- [NTIRE 2024 Challenge on Low Light Image Enhancement: Methods and Results](https://mlsanity.com/paper/2404.14248/) (2024)
- [LYT-NET: Lightweight YUV Transformer-based Network for Low-light Image Enhancement](https://mlsanity.com/paper/2401.15204/) (2024)

## Built on by (most cited first, 5 of 5)

- [LL-GaussianImage: Efficient Image Representation for Zero-shot Low-Light Enhancement with 2D Gaussian Splatting](https://mlsanity.com/paper/2601.15772/) (2026)
- [DST-Net: A Dual-Stream Transformer with Illumination-Independent Feature Guidance and Multi-Scale Spatial Convolution for Low-Light Image Enhancement](https://mlsanity.com/paper/2603.16482/) (2026)
- [Continuous Splatting meets Retinex: Continuous Gaussian Splatting and Implicit Reflectance Modeling for Low-Light Image Enhancement](https://mlsanity.com/paper/2606.16159/) (2026)
- [Gaussian Light Field Splatting: A Physical Prior-Driven Vision Transformer for Unsupervised Low-Light Image Enhancement](https://mlsanity.com/paper/2606.17985/) (2026)
- [AIGS-Net: Compact Illumination Field Modeling via 2D Gaussian Splatting for Fast Low-Light Image Enhancement](https://mlsanity.com/paper/2606.17998/) (2026)

## Abstract (arXiv)

Addressing the urgent need for high-performance real-time low-light image enhancement on resource-constrained edge devices in low-illumination scenarios such as nighttime and tunnels, this paper presents UltraFast-LiNET, an ultra-lightweight network designed for extreme efficiency. The proposed network introduces Dynamic Shift Convolution, DSConv, a highly compact operation with only 12 learnable parameters. By using DSConv to construct a Multi-Scale Shift Residual Block, MSRB, UltraFast-LiNET effectively enlarges the receptive field while maintaining an exceptionally small parameter count. To alleviate gradient instability during the training of lightweight models, we further propose a multi-level gradient-aware loss function that strengthens supervision signals across different feature levels. UltraFast-LiNET supports flexible parameter configurations, enabling the model size to be reduced to as few as 36 learnable parameters. Remarkably, with only 180 learnable parameters, UltraFast-LiNET achieves a PSNR of 19.81 dB on the LOL dataset, surpassing the current state-of-the-art by 0.08 dB. Experimental results demonstrate that the proposed method achieves an effective balance between millisecond-level real-time processing and superior enhancement quality under extreme computational constraints. These results highlight its chip-level computational efficiency and broad prospects for edge-side deployment.