# Bilevel Layer-Positioning LoRA for Real Image Dehazing

arXiv: https://arxiv.org/abs/2603.10872 · HTML: https://mlsanity.com/paper/2603.10872/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Builds on 28 of its 47 references that are in the map.
- First posted on arXiv 11 Mar 2026.

## Builds on (most cited first, 28 of 28)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [MUSIQ: Multi-scale Image Quality Transformer](https://mlsanity.com/paper/2108.05997/) (2021)
- [Benchmarking Single Image Dehazing and Beyond](https://mlsanity.com/paper/1712.04143/) (2017)
- [Multi-Scale Boosted Dehazing Network with Dense Feature Fusion](https://mlsanity.com/paper/2004.13388/) (2020)
- [Domain Adaptation for Image Dehazing](https://mlsanity.com/paper/2005.04668/) (2020)
- [Dense Haze: A benchmark for image dehazing with dense-haze and haze-free images](https://mlsanity.com/paper/1904.02904/) (2019)
- [Iterative Prompt Learning for Unsupervised Backlit Image Enhancement](https://mlsanity.com/paper/2303.17569/) (2023)
- [Pixel-level and Semantic-level Adjustable Super-resolution: A Dual-LoRA Approach](https://mlsanity.com/paper/2412.03017/) (2024)
- [O-HAZE: a dehazing benchmark with real hazy and haze-free outdoor images](https://mlsanity.com/paper/1804.05101/) (2018)
- [DEA-Net: Single image dehazing based on detail-enhanced convolution and content-guided attention](https://mlsanity.com/paper/2301.04805/) (2023)
- [Curricular Contrastive Regularization for Physics-aware Single Image Dehazing](https://mlsanity.com/paper/2303.14218/) (2023)
- [Complexity Experts are Task-Discriminative Learners for Any Image Restoration](https://mlsanity.com/paper/2411.18466/) (2024)
- [RIDCP: Revitalizing Real Image Dehazing via High-Quality Codebook Priors](https://mlsanity.com/paper/2304.03994/) (2023)
- [Selective Hourglass Mapping for Universal Image Restoration Based on Diffusion Model](https://mlsanity.com/paper/2403.11157/) (2024)
- [Convolution Meets LoRA: Parameter Efficient Finetuning for Segment Anything Model](https://mlsanity.com/paper/2401.17868/) (2024)
- [From Synthetic to Real: Image Dehazing Collaborating with Unlabeled Real Data](https://mlsanity.com/paper/2108.02934/) (2021)
- [Nighttime Dehazing with a Synthetic Benchmark](https://mlsanity.com/paper/2008.03864/) (2020)
- [FoundIR: Unleashing Million-scale Training Data to Advance Foundation Models for Image Restoration](https://mlsanity.com/paper/2412.01427/) (2024)
- [UCL-Dehaze: Towards Real-world Image Dehazing via Unsupervised Contrastive Learning](https://mlsanity.com/paper/2205.01871/) (2022)
- [Real-world Image Dehazing with Coherence-based Pseudo Labeling and Cooperative Unfolding Network](https://mlsanity.com/paper/2406.07966/) (2024)
- [Learning a No-Reference Quality Assessment Model of Enhanced Images With Big Data](https://mlsanity.com/paper/1904.08632/) (2019)
- [Iterative Predictor-Critic Code Decoding for Real-World Image Dehazing](https://mlsanity.com/paper/2503.13147/) (2025)
- [HazyDet: Open-Source Benchmark for Drone-View Object Detection with Depth-Cues in Hazy Scenes](https://mlsanity.com/paper/2409.19833/) (2024)
- [Prompt-based test-time real image dehazing: a novel pipeline](https://mlsanity.com/paper/2309.17389/) (2023)
- [Task-Oriented Convex Bilevel Optimization with Latent Feasibility](https://mlsanity.com/paper/1907.03083/) (2019)
- [CoA: Towards Real Image Dehazing via Compression-and-Adaptation](https://mlsanity.com/paper/2504.05590/) (2025)
- [PHATNet: A Physics-guided Haze Transfer Network for Domain-adaptive Real-world Image Dehazing](https://mlsanity.com/paper/2507.14826/) (2025)
- [Contribution-based Low-Rank Adaptation with Pre-training Model for Real Image Restoration](https://mlsanity.com/paper/2408.01099/) (2024)

## Built on by (most cited first, 6 of 6)

- [PRISM: Rethinking Atmospheric Scattering Reconstruction as a Unified Understanding and Restoration Model for Real-world Dehazing](https://mlsanity.com/paper/2604.07048/) (2026)
- [Robust-U1: Can MLLMs Self-Recover Corrupted Visual Content for Robust Understanding?](https://mlsanity.com/paper/2606.08063/) (2026)
- [Towards UAV Image Dehazing: A UAV Atmospheric Scattering Model, Benchmark, and Geometry-Aware Deep Unfolding Network](https://mlsanity.com/paper/2606.16392/) (2026)
- [Efficient Real-World Dehazing via Physics-Inspired Global-Local Decoupling](https://mlsanity.com/paper/2606.25732/) (2026)
- [HazeSpikeMamba: Coupling Spiking-Inspired and State-Space Features for Self-Supervised Real-World Dehazing](https://mlsanity.com/paper/2608.06886/) (2026)
- [Tri-DehazeGS: Scene--Medium Decoupled Gaussian Splatting with Transmittance-Aware Optimization](https://mlsanity.com/paper/2609.11223/) (2026)

## Abstract (arXiv)

Learning-based real image dehazing methods have achieved notable progress, yet they still face adaptation challenges in diverse real haze scenes. These challenges mainly stem from the lack of effective unsupervised mechanisms for unlabeled data and the heavy cost of full model fine-tuning. To address these challenges, we propose the haze-to-clear text-directed loss that leverages CLIP's cross-modal capabilities to reformulate real image dehazing as a semantic alignment problem in latent space, thereby providing explicit unsupervised cross-modal guidance in the absence of reference images. Furthermore, we introduce the Bilevel Layer-positioning LoRA (BiLaLoRA) strategy, which learns both the LoRA parameters and automatically search the injection layers, enabling targeted adaptation of critical network layers. Extensive experiments demonstrate our superiority against state-of-the-art methods on multiple real-world dehazing benchmarks. The code is publicly available at https://github.com/YanZhang-zy/BiLaLoRA.