LCMamNet: A Lightweight Cross-scale Mamba Network for Infrared Small Target Detection
2026-07-27 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors developed a lightweight neural network called LCMamNet to detect very small targets in infrared images, which is hard because these targets are tiny and easily hidden by background noise. Their method enhances the details of these small targets while reducing distractions from the background by using special blocks that focus on directional features, a cross-scale feature fusion step, and a decoder that selects important details. They tested the network on several infrared datasets and achieved good accuracy with low computational cost, making it suitable for real-time use on edge devices like NVIDIA Jetson. The authors also shared their code publicly for others to use.
Infrared small target detectionLightweight neural networkCross-scale feature fusionDirectional bottleneck residual blockMamba networkHierarchical encoderDecodermIoUReal-time edge inferenceNVIDIA Jetson
Authors
Yuhao Fan, Le Hui, Yuchao Dai
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.