Air-Know: Arbiter-Calibrated Knowledge-Internalizing Robust Network for Composed Image Retrieval
Zhiheng Fu Yupeng Hu Qianyun Yang Shiqi Zhang Zhiwei Chen Zixu Li Shandong University {fuzhiheng8,zivczw,lizixu.cs}@gmail.com;{qianyunyang,zhangshiqi}@mail.sdu.edu.cn; [email protected] https://github.com/ZhihFu/Air-Know/ · 2026
Read Air-Know: Arbiter-Calibrated Knowledge-Internalizing Robust Network for Composed Image Retrieval by Zhiheng Fu Yupeng Hu Qianyun Yang Shiqi Zhang Zhiwei Chen Zixu Li Shandong University {fuzhiheng8,zivczw,lizixu.cs}@gmail.com;{qianyunyang,zhangshiqi}@mail.sdu.edu.cn; [email protected] https://github.com/ZhihFu/Air-Know/ (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-19386.txt.
Description
Fu and colleagues propose Air-Know, a network for composed image retrieval that tackles noisy triplet correspondence, where query triplets carry unreliable matches. Most robust learning methods lean on the small-loss hypothesis, but semantic ambiguity in noisy triplets, such as partial matching, invalidates that assumption. It also leads to shaky noise spotting. The learner then falls into a self-fed vicious cycle, tangled with its arbiter, causing catastrophic representation pollution. Air-Know breaks that cycle with an Expert-Proxy-Diversion decoupling design, starting with external prior arbitration.
Catalog record
| CREATOR | Zhiheng Fu Yupeng Hu Qianyun Yang Shiqi Zhang Zhiwei Chen Zixu Li Shandong University {fuzhiheng8,zivczw,lizixu.cs}@gmail.com;{qianyunyang,zhangshiqi}@mail.sdu.edu.cn; [email protected] https://github.com/ZhihFu/Air-Know/ |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 80120 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.19386