CSHNet: A Novel Information Asymmetric Image Translation Method
Xi Yang, Haoyuan Shi, Zihan Wang, Nannan Wang, and · 2025
Read CSHNet: A Novel Information Asymmetric Image Translation Method by Xi Yang, Haoyuan Shi, Zihan Wang, Nannan Wang, and (2025) 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-2501-10197.txt.
Description
Xi Yang, Haoyuan Shi, Zihan Wang, Nannan Wang, and Xinbo Gao present CSHNet in 2025 for translating images across domains with uneven detail. Tasks like radar-to-photo conversion must invent rich content from sparse input. Standard networks catch fine detail but lose global shape, merging regions that should stay apart.
Their hybrid network braids two designs: one strong on detail, one strong on structure. Cross links let the two halves trade information as they work, and a special loss guards edges during conversion. On scene-level and object-level data sets, the method beat prior tools in both looks and scores. Code is shared in public.
Catalog record
| CREATOR | Xi Yang, Haoyuan Shi, Zihan Wang, Nannan Wang, and |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 52 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-10197