Abstract: As the accuracy of salient object detection has been improved, the focus has gradually shifted to how to improve the accuracy of small object detection. However, existing object detection methods mainly study the general object detection problem using visible images as input, and the most of small object detection methods are designed for visible images, leaving the small object detection on infrared images underexplored. Compared with standard scale objects, infrared small objects lack color information, which makes them more dependent on contextual information. To solve this problem, an infrared small object detection model is proposed based on the standard YOLOv5 model. The local information around small objects is effectively combined with global information by the newly introduced Dynamic Contextual Information Extraction Module, which adapts to the subtle morphological changes of infrared small objects dynamically. The Channel-Detail Attention Module is designed to aggregate the channel and detail information of the infrared small objects to improve the accuracy of the regression. To use the shallow features to enhance the detail information of infrared small objects, the feature with new scales are upsampled and fused with the shallow features, which captures more detailed information about infrared small objects and avoids feature blending. To demonstrate the effectiveness of the proposed method, extensive experiments are conducted on the public infrared datasets, including ITTD, IRSTD-1k, and NUAA-SIRST. The experimental results show that the proposed method outperforms the compared methods in terms of mAP by 5.1% on the ITTD dataset, and the mAP is also improved by 3.7% when compared to the baseline method (i.e., YOLOv5). Results on the IRSTD-1k and NUAA-SIRST datasets also demonstrate the effectiveness of our design. An ablation study is performed to verify the effectiveness of different modules. The proposed infrared small object detection model is robust to the infrared small objects in complex backgrounds, which improves the accuracy and reduces the false alarm rate of infrared small object detection effectively.

Key words: Deep learning, Object detection, Infrared small object detection, Attention mechanism, YOLOv5