Abstract:Limited by equipment limitations and the lack of target intrinsic features, existing infrared small target detection methods have difficulty meeting actual comprehensive performance requirements. Therefore, we propose an innovative lightweight and robust network (LR-Net), which abandons the complex structure and achieves an effective balance between detection accuracy and resource consumption. Specifically, to ensure the lightweight and robustness, on the one hand, we construct a lightweight feature extraction attention (LFEA) module, which can fully extract target features and strengthen information interaction across channels. On the other hand, we construct a simple refined feature transfer (RFT) module. Compared with direct cross-layer connections, the RFT module can improve the network's feature refinement extraction capability with little resource consumption. Meanwhile, to solve the problem of small target loss in high-level feature maps, on the one hand, we propose a low-level feature distribution (LFD) strategy to use low-level features to supplement the information of high-level features. On the other hand, we introduce an efficient simplified bilinear interpolation attention module (SBAM) to promote the guidance constraints of low-level features on high-level features and the fusion of the two. In addition, We abandon the traditional resizing method and adopt a new training and inference cropping strategy, which is more robust to datasets with multi-scale samples. Extensive experimental results show that our LR-Net achieves state-of-the-art (SOTA) performance. Notably, on the basis of the proposed LR-Net, we achieve 3rd place in the "ICPR 2024 Resource-Limited Infrared Small Target Detection Challenge Track 2: Lightweight Infrared Small Target Detection".

RLPGB-Net: Reinforcement Learning of Feature Fusion and Global Context Boundary Attention for Infrared Dim Small Target Detection

MRF3Net: Infrared Small Target Detection Using Multi-Receptive Field Perception and Effective Feature Fusion

Local Information Guided Global Integration for Infrared Small Target Detection.

An Infrared Target Intrusion Detection Method Based on Feature Fusion and Enhancement

A Nonlocal Enhanced Feature Pyramid Network for Infrared Small Target Detection

Global Induced Local Network for Infrared Dim Small Target Detection

LRAF-Net: Long-Range Attention Fusion Network for Visible-Infrared Object Detection.

MAGPNet: Multi-Domain Attention-Guided Pyramid Network for Infrared Small Object Detection

DSDANet: Infrared Dim Small Target Detection Via Attention Enhanced Feature Fusion Network

Local Patch Network with Global Attention for Infrared Small Target Detection

Cross-Connected Bidirectional Pyramid Network for Infrared Small-Dim Target Detection

Context-aware Cross-Level Attention Fusion Network for Infrared Small Target Detection

Multi-scale feature fusion attention network for infrared small target detection

SFFNet: Shallow Feature Fusion Network Based on Detection Framework for Infrared Small Target Detection

Edge-Guided Perceptual Network for Infrared Small Target Detection

AGPCNet: Attention-Guided Pyramid Context Networks for Infrared Small Target Detection

Infrared Small Target Detection Using Focally Multi-Patch Network

Infrared Dim and Small Target Detection Based on Local–Global Feature Fusion

AFE-Net: Attention-Guided Feature Enhancement Network for Infrared Small Target Detection

Fine-Grained Guided Model Fusion Network with Attention Mechanism for Infrared Small Target Segmentation

LR-Net: A Lightweight and Robust Network for Infrared Small Target Detection