Diffusion-Based Continuous Feature Representation for Infrared Small-Dim Target Detection

Linyu Fan, Yingying Wang, Guoliang Hu, Feifei Li, Yuhang Dong, Hui Zheng, Changqing Lin, Yue Huang, Xinghao Ding

Abstract: Infrared small-dim target detection (ISDTD) plays a pivotal role in missions involving rescue, surveillance, and early warning systems. Despite remarkable strides made by existing methods, certain limitations still hinder the detection accuracy, including deficiency in high-resolution (HR) representation, inadequacy in addressing dim targets, and difficulty in tackling low-contrast targets against complex backgrounds. To overcome these limitations, we propose a diffusion-based continuous feature representation network (DCFR-Net), comprising two crucial branches: diffusion-based continuous HR feature representation (DCHFR) and ISDTD. Specifically, to precisely capture extremely small target contours, DCHFR integrates implicit neural representation (INR) into a conditional denoising diffusion model, super-resolving infrared targets in a self-supervised strategy. ISDTD leverages the shared encoder from DCHFR to construct HR feature representation, which is fed into multiscale implicit feature alignment (MIFA) and spatial-frequency feature interaction (SFFI). To alleviate the impact of dim and vulnerable targets, MIFA delicately aggregates different-layer features in a resolution-free manner. Furthermore, to enhance the contrast between infrared targets and intricate backgrounds, SFFI achieves profound SFFI and global-local receptive field mixture. Extensive experiments conducted on three challenging datasets of NUAA-SIRST, IRSTD-1k, and NUDT-SIRST reveal that our DCFR-Net outperforms the state-of-the-art (SOTA) methods, demonstrating the superiority and robustness of our approach in ISDTD. The code will be available at https://github.com/flyannie/DCFR-Net .