GS-YOLO: A lightweight high-accuracy model for small target detection in drone aerial images

Xiaoyuan Jin, Xiyuan Zhu, Dongdong Kang, Wangyu Shen, Yang Zhao, Xun Li, Baoxi Yuan, Yuzhen Zhao · 2026

Read GS-YOLO: A lightweight high-accuracy model for small target detection in drone aerial images by Xiaoyuan Jin, Xiyuan Zhu, Dongdong Kang, Wangyu Shen, Yang Zhao, Xun Li, Baoxi Yuan, Yuzhen Zhao (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/plos-pone-0350840.txt.

Description

Xiaoyuan Jin and colleagues present an ultra light detector for small targets in drone photos, where tiny objects drown in clutter and scale swings. Their design swaps heavy maths blocks for lean ghost style ones, adds a two path attention step that damps background noise, and focuses channel weights on faint small target cues.

On a public drone photo set, accuracy nudged above a strong compact rival while the model held under one million settings. That tiny size suits on board drone chips with strict power limits. Traffic counts, crowd checks, and small vehicle spotting from the air all need exactly this mix of lightness and sharp small target vision.

Catalog record

CREATORXiaoyuan Jin, Xiyuan Zhu, Dongdong Kang, Wangyu Shen, Yang Zhao, Xun Li, Baoxi Yuan, Yuzhen Zhao
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE48443
LICENSECC_BY
RIGHTSCC BY
SOURCEPLOS

Source: https://doi.org/10.1371/journal.pone.0350840

Browse the full library · About the Tower of Babel