DAN-YOLO-L: a lightweight YOLO approach for efficient infrared small object detection from UAV perspectives
Kai Feng, Guojun Lin, Tong Lin, Zhongqiang Luo, Run Ma · 2026
Read DAN-YOLO-L: a lightweight YOLO approach for efficient infrared small object detection from UAV perspectives by Kai Feng, Guojun Lin, Tong Lin, Zhongqiang Luo, Run Ma (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/PMC13575193.txt.
Description
Kai Feng and four colleagues present a lighter detection model for finding small warm objects in drone infrared images, in a 2026 paper. Tiny targets, messy backgrounds, and faint features make this task hard, and drone hardware needs fast, small models. Their design, DAN-YOLO-L, reworks the YOLO detection framework.
A new feature module sharpens detail across scales, an auxiliary head improves how features at different levels interact, and a revised loss function steadies box placement for tiny targets. Pruning then cut the parameter count by 20.8 percent while speed rose by 8.32 percent. On the HIT-UAV benchmark the model scored 81.9 percent on the main detection measure. A second infrared dataset confirmed the gains carried over.
Catalog record
| CREATOR | Kai Feng, Guojun Lin, Tong Lin, Zhongqiang Luo, Run Ma |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 57779 |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |