CAFF-Net-method-and-the-NCHU-SIRST-dataset

CAFF-Net method and the NCHU-SIRST dataset

Paper

An Infrared Small Target Detection Method Using Coordinate Attention and Feature Fusion

Qi Shi1, Congxuan Zhang1,2, Zhen Chen1, Feng Lu1, Liyue Ge3, Shuigen Wei3

1 School of Measuring and Optical Engineering, Nanchang Hangkong University, Nanchang 330063, China;

2 Institute of Automation, Chinese Academy of Sciences, Beijing 100190, China;

3 School of Information Engineering, Nanchang Hangkong University, Nanchang 330063, China;

This paper has been submitted to Infrared Physics & Technology

Dataset Description

The NCHU-SIRST dataset are divided into 273 training frames and 317 test frames, and the target scene is roughly classified into six categories: architecture, cloudless sky, complex clouds, continuous clouds, sea, and trees.The annotation form is XML.

Dataset Statistics