Pseudo-Sample Generation and Self-Supervised Framework for Infrared Dim and Small Target Detection
Jinxin Guo, Weida Zhan, Dehua Huo, Depeng Zhu, Yu Chen, Yichun Jiang, Xiaoyu Xu, Jun Chen · 2025
Read Pseudo-Sample Generation and Self-Supervised Framework for Infrared Dim and Small Target Detection by Jinxin Guo, Weida Zhan, Dehua Huo, Depeng Zhu, Yu Chen, Yichun Jiang, Xiaoyu Xu, Jun Chen (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/pmc-12732119.txt.
Description
Infrared cameras hunting a dim, tiny target have almost no labeled examples to learn from, and fake ones made by pasting patches ignore how real sensors degrade an image. Jinxin Guo and co-authors generate training data from physics instead, modeling how targets and backgrounds each decay through an infrared system, then constraining the fakes by global, local, and brightness statistics.
A self-supervised detector framework trains on the synthetic set with custom losses and two-dimensional quality metrics. In their tests the physics-made data beat existing simulated sets on realism and lifted detection. The work treats data generation itself as an optics problem.
Catalog record
| CREATOR | Jinxin Guo, Weida Zhan, Dehua Huo, Depeng Zhu, Yu Chen, Yichun Jiang, Xiaoyu Xu, Jun Chen |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 72 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |