| title | Robust-U1: Can MLLMs Self-Recover Corrupted Visual Content for Robust Understanding? |
|---|---|
| authors | Jiaqi Tang, Jianmin Chen, Youyang Zhai, Wei Wei, Runtao Liu, Mengjie Zhao, Xiangyu Wu, Qingfa Xiao, Qifeng Chen |
| date | 2026-06-06 |
| venue | arXiv (preprint, not peer reviewed) |
| id | arXiv:2606.08063 |
| upvotes | 32 (Hugging Face Daily Papers, 2026-11-01) |
| links | abs pdf hf code |
| abstract | Multimodal Large Language Models (MLLMs) have demonstrated remarkable success in visual understanding, yet their performance degrades significantly under real-world visual corruptions. While existing robustness enhancement approaches exist, they are limited: black-box feature alignment lacks interpretability, and white-box text-based reasoning cannot restore lost pixel-level details. This work investigates a fundamental research question: Can MLLMs recover corrupted visual content by themselves? To address this, we propose Robust-U1, a novel framework that equips MLLMs with explicit visual self-recovery capability for robust understanding. The approach comprises three core stages: supervised fine-tuning for initial reconstruction, reinforcement learning with dual rewards (pixel-level SSIM and semantic-level CLIP similarity) for aligning high visual quality, and multimodal reasoning that jointly considers both the corrupted input and the recovered image. Extensive experiments demonstrate that Robust-U1 achieves state-of-the-art robustness on the real-world corruption benchmark and maintains superior performance under adversarial corruptions on general VQA benchmarks. Analysis confirms that high-quality visual recovery directly enhances reasoning performance, establishing self-recovery as a critical mechanism for robust visual understanding. The source code is available at https://github.com/jqtangust/Robust-U1. |
cite
@misc{tang2026robust,
title={Robust-U1: Can MLLMs Self-Recover Corrupted Visual Content for Robust Understanding?},
author={Jiaqi Tang and Jianmin Chen and Youyang Zhai and Wei Wei and Runtao Liu and Mengjie Zhao and Xiangyu Wu and Qingfa Xiao and Qifeng Chen},
year={2026},
eprint={2606.08063},
archivePrefix={arXiv},
url={https://arxiv.org/abs/2606.08063}
}