# Dynamic Pre-training: Towards Efficient and Scalable All-in-One Image Restoration

arXiv: https://arxiv.org/abs/2404.02154 · HTML: https://mlsanity.com/paper/2404.02154/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- Builds on 36 of its 62 references that are in the map.
- First posted on arXiv 2 Apr 2024.

## Builds on (most cited first, 30 of 36)

- [Masked Autoencoders Are Scalable Vision Learners](https://mlsanity.com/paper/2111.06377/) (2021)
- [LAION-5B: An open large-scale dataset for training next generation image-text models](https://mlsanity.com/paper/2210.08402/) (2022)
- [Restormer: Efficient Transformer for High-Resolution Image Restoration](https://mlsanity.com/paper/2111.09881/) (2021)
- [Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising](https://mlsanity.com/paper/1608.03981/) (2016)
- [Pre-Trained Image Processing Transformer](https://mlsanity.com/paper/2012.00364/) (2020)
- [Uformer: A General U-Shaped Transformer for Image Restoration](https://mlsanity.com/paper/2106.03106/) (2021)
- [Multi-Stage Progressive Image Restoration](https://mlsanity.com/paper/2102.02808/) (2021)
- [Simple Baselines for Image Restoration](https://mlsanity.com/paper/2204.04676/) (2022)
- [ImageNet-21K Pretraining for the Masses](https://mlsanity.com/paper/2104.10972/) (2021)
- [FFDNet: Toward a Fast and Flexible Solution for CNN based Image Denoising](https://mlsanity.com/paper/1710.04026/) (2017)
- [Benchmarking Single Image Dehazing and Beyond](https://mlsanity.com/paper/1712.04143/) (2017)
- [Learning Deep CNN Denoiser Prior for Image Restoration](https://mlsanity.com/paper/1704.03264/) (2017)
- [NIMA: Neural Image Assessment](https://mlsanity.com/paper/1709.05424/) (2017)
- [Learning Texture Transformer Network for Image Super-Resolution](https://mlsanity.com/paper/2006.04139/) (2020)
- [DehazeNet: An End-to-End System for Single Image Haze Removal](https://mlsanity.com/paper/1601.07661/) (2016)
- [Learning Enriched Features for Real Image Restoration and Enhancement](https://mlsanity.com/paper/2003.06792/) (2020)
- [MAXIM: Multi-Axis MLP for Image Processing](https://mlsanity.com/paper/2201.02973/) (2022)
- [Multi-Scale Progressive Fusion Network for Single Image Deraining](https://mlsanity.com/paper/2003.10985/) (2020)
- [Density-aware Single Image De-raining using a Multi-stream Dense Network](https://mlsanity.com/paper/1802.07412/) (2018)
- [Multi-Scale Boosted Dehazing Network with Dense Feature Fusion](https://mlsanity.com/paper/2004.13388/) (2020)
- [CycleISP: Real Image Restoration via Improved Data Synthesis](https://mlsanity.com/paper/2003.07761/) (2020)
- [Deblurring by Realistic Blurring](https://mlsanity.com/paper/2004.01860/) (2020)
- [Uncertainty Guided Multi-Scale Residual Learning-using a Cycle Spinning CNN for Single Image De-Raining](https://mlsanity.com/paper/1906.11129/) (2019)
- [PromptIR: Prompting for All-in-One Blind Image Restoration](https://mlsanity.com/paper/2306.13090/) (2023)
- [Exploring the Limits of Large Scale Pre-training](https://mlsanity.com/paper/2110.02095/) (2021)
- [FD-GAN: Generative Adversarial Networks with Fusion-discriminator for Single Image Dehazing](https://mlsanity.com/paper/2001.06968/) (2020)
- [A General Decoupled Learning Framework for Parameterized Image Operators](https://mlsanity.com/paper/1907.05852/) (2019)
- [Burst Image Restoration and Enhancement](https://mlsanity.com/paper/2110.03680/) (2021)
- [Residual Local Feature Network for Efficient Super-Resolution](https://mlsanity.com/paper/2205.07514/) (2022)
- [BANet: Blur-aware Attention Networks for Dynamic Scene Deblurring](https://mlsanity.com/paper/2101.07518/) (2021)

## Built on by (most cited first, 8 of 8)

- [A Survey on All-in-One Image Restoration: Taxonomy, Evaluation and Future Trends](https://mlsanity.com/paper/2410.15067/) (2024)
- [HAIR: Hypernetworks-based All-in-One Image Restoration](https://mlsanity.com/paper/2408.08091/) (2024)
- [Efficient Degradation-aware Any Image Restoration](https://mlsanity.com/paper/2405.15475/) (2024)
- [Efficient Degradation-agnostic Image Restoration via Channel-Wise Functional Decomposition and Manifold Regularization](https://mlsanity.com/paper/2505.18679/) (2025)
- [Any Image Restoration via Efficient Spatial-Frequency Degradation Adaptation](https://mlsanity.com/paper/2504.14249/) (2025)
- [Clear Roads, Clear Vision: Advancements in Multi-Weather Restoration for Smart Transportation](https://mlsanity.com/paper/2510.09228/) (2025)
- [DiffuMural: Restoring Dunhuang Murals with Multi-scale Diffusion](https://mlsanity.com/paper/2504.09513/) (2025)
- [Pruning Overparameterized Multi-Task Networks for Degraded Web Image Restoration](https://mlsanity.com/paper/2510.14463/) (2025)

## Abstract (arXiv)

All-in-one image restoration tackles different types of degradations with a unified model instead of having task-specific, non-generic models for each degradation. The requirement to tackle multiple degradations using the same model can lead to high-complexity designs with fixed configuration that lack the adaptability to more efficient alternatives. We propose DyNet, a dynamic family of networks designed in an encoder-decoder style for all-in-one image restoration tasks. Our DyNet can seamlessly switch between its bulkier and lightweight variants, thereby offering flexibility for efficient model deployment with a single round of training. This seamless switching is enabled by our weights-sharing mechanism, forming the core of our architecture and facilitating the reuse of initialized module weights. Further, to establish robust weights initialization, we introduce a dynamic pre-training strategy that trains variants of the proposed DyNet concurrently, thereby achieving a 50% reduction in GPU hours. Our dynamic pre-training strategy eliminates the need for maintaining separate checkpoints for each variant, as all models share a common set of checkpoints, varying only in model depth. This efficient strategy significantly reduces storage overhead and enhances adaptability. To tackle the unavailability of large-scale dataset required in pre-training, we curate a high-quality, high-resolution image dataset named Million-IRD, having 2M image samples. We validate our DyNet for image denoising, deraining, and dehazing in all-in-one setting, achieving state-of-the-art results with 31.34% reduction in GFlops and a 56.75% reduction in parameters compared to baseline models. The source codes and trained models are available at https://github.com/akshaydudhane16/DyNet.