LCMamNet: A Lightweight Cross-scale Mamba Network for Infrared Small Target Detection

2026-07-27

AI Summary

This work addresses the challenges of infrared small target detection, where targets exhibit sparse pixel occupancy and are highly susceptible to interference from cloud clutter and background noise, while existing lightweight methods struggle to simultaneously preserve local structural details and suppress background. To this end, we propose LCMamNet, which introduces a Cross-shaped Directional Bottleneck Residual (CDBR) module to enhance directional sensitivity with minimal computational overhead, and designs a Latent-space Dense Cross-scale Fusion (LDCF) mechanism coupled with bidirectional Mamba for efficient hierarchical semantic reorganization. A progressive decoder further enables effective background suppression without compromising fine details. The model achieves mIoU scores of 71.25%, 79.60%, and 95.58% on IRSTD-1k, NUAA-SIRST, and NUDT-SIRST, respectively, with only 1.175M parameters and 6.91 GFLOPs, and demonstrates real-time performance with an average inference latency of 6.62 ms, successfully deployed on Jetson Orin NX.

Abstract

Infrared small target detection (IRSTD) is important for low-altitude perception, unmanned-system warning, and security monitoring. However, weak targets in infrared imagery usually occupy only a few pixels and are easily submerged by cloud clutter, ground edges, and bright noise, making it difficult for lightweight segmentation-based methods to preserve local target structures while suppressing background interference. To address these challenges, we propose LCMamNet, a lightweight cross-scale Mamba network that progressively enhances local target structures, interacts cross-scale context in a latent space, and restores spatial details with background suppression. Specifically, a compact hierarchical encoder with cross-shaped directional bottleneck residual (CDBR) blocks strengthens direction-sensitive target structures under a small computation budget. A latent dense cross-scale fusion (LDCF) module then performs dense all-level interaction through bidirectional Mamba modeling and reorganizes the interacted features into stable hierarchical semantics. Finally, a progressive decoder selectively recovers shallow spatial details while suppressing irrelevant background textures. Extensive experiments on IRSTD-1k, NUAA-SIRST, and NUDT-SIRST show that the proposed network achieves mIoU scores of 71.25\%, 79.60\%, and 95.58\%, respectively, with only 1.175M parameters and 6.91 GFLOPs. It also runs with a mean inference latency of 6.62 ms, and deployment results on an NVIDIA Jetson Orin NX 16G SUPER further demonstrate its practical potential for real-time edge inference. The code and checkpoints are publicly available at https://github.com/Haoyu096/LCMamNet.

Problem

• Infrared small target detection
• background suppression
• local target structure
• lightweight segmentation
• cloud clutter

Innovation

• Mamba
• cross-scale fusion
• lightweight network
• infrared small target detection
• latent space interaction