RSFNet: A retention-based network with spatial–frequency joint enhancement for infrared small target detection

Zhicheng Tan, Shanlin Sun, Guo Li · 2026

Read RSFNet: A retention-based network with spatial–frequency joint enhancement for infrared small target detection by Zhicheng Tan, Shanlin Sun, Guo Li (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-0345722.txt.

Description

Zhicheng Tan, Shanlin Sun, and Guo Li present in 2026 a neural network for spotting tiny hot targets in infrared images, such as distant ships or vehicles. Such targets give weak signals and little texture, so they are easy to miss or confuse with background noise. Their network uses a retention style attention that links far apart parts of an image without blurring fine detail.

An added module joins space and frequency views to pass clear target features through the system. Tests on several public datasets beat leading methods and cut training time by nearly a third. Faster, sharper spotting helps watching traffic and coasts from a distance.

Catalog record

CREATORZhicheng Tan, Shanlin Sun, Guo Li
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE34430
LICENSECC_BY
RIGHTSCC BY
SOURCEPLOS

Source: https://doi.org/10.1371/journal.pone.0345722

Browse the full library · About the Tower of Babel