Infrared small target detection (ISTD) means distinguishing small and faint targets from IR images. Small targets typically span only a handful of pixels, lacking distinct texture and clear structural details. For the past few years, deep learning has made big strides in the field of ISTD. Yet, a persistent challenge remains: the lack of high-level semantic information may cause the disappearance of small target features in the network’s deep layers, ultimately impairing detection accuracy. To tackle this problem, we introduce WT-HMFF, an innovative network architecture that combines the Wavelet Transform Convolution (WTConv) module with the Hierarchical Multi-Scale Feature Fusion (HMFF) module to enhance the ISTD algorithm’s performance. WTConv expands the receptive field through wavelet convolution, effectively capturing global contextual information while preserving target shape characteristics. The HMFF module enables the efficient fusion of shallow and deep features, maintaining the high resolution of deep feature maps and preventing the disappearance of small target features. We have tested it out on public datasets, SIRST and IRSTD-1k, and validated the superiority and robustness of WT-HMFF compared to other methods.

Infrared small target detection (ISTD) plays a crucial role in many real-world applications. However, this task remains highly challenging due to the extremely small target size, low contrast, and complex background interference as infrared small targets often occupy fewer than 80 pixels in a $256\times 256$ image under a commonly used ISTD criterion. Although Segment Anything Model (SAM) shows strong generalization in image segmentation, directly applying SAM to ISTD is suboptimal, primarily due to the significant modality gap between RGB and infrared imagery, as well as the prohibitive cost of full-parameter fine-tuning. To address these challenges, we propose a prompt-free and parameter-efficient fine-tuning framework that adapts SAM for ISTD. To bridge the cross-modality gap while preserving the pretrained prior knowledge of SAM, a lightweight Infrared Adapter (IR-Adapter) is introduced into the image encoder, enabling effective task adaptation with only a small number of trainable parameters. Furthermore, to alleviate the loss of small target information in deep network layers, we design a Multi-Scale Feature Fusion (MSF) module that integrates hierarchical features from different encoder stages. In addition, a Coarse-to-Fine Head (CFH) with dual-branch prediction is proposed to incorporate fine-grained details for more accurate target localization and segmentation. Extensive experiments conducted on two public datasets demonstrate that the proposed method achieves better overall performance than existing representative approaches, yielding higher IoU, nIoU and Pd.