The research team led by Professor Xu Tingfa from the School of Optoelectronics at Beijing Institute of Technology (BIT) has made new progress in infrared small cluster target detection.

Addressing core challenges such as insufficient utilization of temporal information in infrared imaging, low signal-to-noise ratio of small cluster targets, and complex backgrounds, the team proposed a one-step transformer detection method.

Their research findings, titled OSFormer: One-Step Transformer for Infrared Video Small Object Detection, have been published in the top-tier international journal in the field of computer vision and image processing, IEEE Transactions on Image Processing (TIP, CCF-A). TIP is a leading academic journal in the field of image processing under the IEEE and is rated as a Q1 top journal by the Chinese Academy of Sciences SCI classification. It is also recommended as an A-level journal by the China Computer Federation.

OSFormer diagram

Through validation on different infrared datasets, the proposed one-step transformer detection paradigm (OSFormer) structure has achieved accurate detection of single and clustered low-altitude drone targets in complex backgrounds such as urban and mountainous areas.

Paper details：Haolin Qin, Tingfa Xu, Yuan Tang, Fengxiang Xu and Jianan Li. 2025. OSFormer: One-Step Transformer for Infrared Video Small Object Detection. IEEE Transactions on Image Processing, doi: 10.1109/TIP.2025.3598426.