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".

Global and Local Attention-based Inception U-Net for Static IR Drop Estimation

Global and Local Attention-based Inception U-Net for Static IR Drop Prediction

Static IR Drop Prediction with Attention U-Net and Saliency-Based Explainability

Attention Res-Unet: an Efficient Shadow Detection Algorithm

PowerNet: Transferable Dynamic IR Drop Estimation via Maximum Convolutional Neural Network

PDNNet: PDN-Aware GNN-CNN Heterogeneous Network for Dynamic IR Drop Prediction

Circuits Physics Constrained Predictor of Static IR Drop with Limited Data

Fast Analysis of Static IR Drop Effect Based on Machine Learning Methods

Dense Nested Attention Network for Infrared Small Target Detection

FDDBA-NET: Frequency Domain Decoupling Bidirectional Interactive Attention Network for Infrared Small Target Detection

Fast Dynamic IR-Drop Prediction Using Machine Learning in Bulk FinFET Technologies

UCDnet: Double U-Shaped Segmentation Network Cascade Centroid Map Prediction for Infrared Weak Small Target Detection

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

Paying more attention to local contrast: improving infrared small target detection performance via prior knowledge

A Lightweight Inception Boosted U-Net Neural Network for Routability Prediction

Thermal and IR Drop Analysis Using Convolutional Encoder-Decoder Networks

Local Patch Network with Global Attention for Infrared Small Target Detection

Infrared target detection algorithm based on multipath coordinate attention mechanism

Global Induced Local Network for Infrared Dim Small Target Detection

Location-Guided Dense Nested Attention Network for Infrared Small Target Detection

Lost in UNet: Improving Infrared Small Target Detection by Underappreciated Local Features