HTTP/2 200 server: GitHub.com content-type: text/html; charset=utf-8 last-modified: Sat, 11 Oct 2025 11:45:43 GMT access-control-allow-origin: * strict-transport-security: max-age=31556952 etag: W/"68ea4367-89ef" expires: Mon, 29 Dec 2025 18:04:31 GMT cache-control: max-age=600 content-encoding: gzip x-proxy-cache: MISS x-github-request-id: 2919:2D64E0:91DBF9:A3AE6C:6952C057 accept-ranges: bytes age: 0 date: Mon, 29 Dec 2025 17:54:32 GMT via: 1.1 varnish x-served-by: cache-bom-vanm7210098-BOM x-cache: MISS x-cache-hits: 0 x-timer: S1767030872.844239,VS0,VE202 vary: Accept-Encoding x-fastly-request-id: c5351b358b268adf64b05bd992a40028a17d923e content-length: 5866

4 K A g e n t :

Agentic Any Image to 4K Super-Resolution

Yushen Zuo¹, Qi Zheng¹⁺, Mingyang Wu¹⁺, Xinrui Jiang²⁺, Renjie Li¹,
Jian Wang³, Yide Zhang⁴, Gengchen Mai⁵, Lihong V. Wang⁶, James Zou²,
Xiaoyu Wang⁷, Ming-Hsuan Yang⁸, Zhengzhong Tu^1*

¹Texas A&M University    ²Stanford University    ³Snap Inc.    ⁴CU Boulder
   ⁵UT Austin    ⁶California Institute of Technology    ⁷Topaz Labs    ⁸UC Merced
⁺Indicates Equal Contribution
^*Corresponding Author

Accepted by NeurIPS 2025

Paper Code

We present 4KAgent, an agentic image super-resolution generalist designed to universally upscale any image to 4K resolution, regardless of input type, degradation level, or domain. That is, 4KAgent effectively restores diverse imagery, spanning from natural scenes, severely degraded captures (e.g., old photos), human/pet portraits, AI-generated content (AIGC), as well as specialized scientific imaging domains, such as remote sensing, fluorescence microscopy, pathology, and various medical modalities like X-ray, ultrasound, and funduscopy --- all without the need for any re-training or domain-specific adaptation.

4KAgent Introduction Video

Pipeline

Gallery

(Please refresh the page if not display normally.)

Natural Images

AIGC Images

Scientific Imagery

BibTeX

@article{zuo20254kagent,
        title={4KAgent: Agentic Any Image to 4K Super-Resolution}, 
        author={Yushen Zuo and Qi Zheng and Mingyang Wu and Xinrui Jiang and Renjie Li and Jian Wang and Yide Zhang and Gengchen Mai and Lihong V. Wang and James Zou and Xiaoyu Wang and Ming-Hsuan Yang and Zhengzhong Tu},
        year={2025},
        eprint={2507.07105},
        archivePrefix={arXiv},
        primaryClass={cs.CV},
        url={https://arxiv.org/abs/2507.07105}, 
}