MLSDNet: Multiclass Lightweight SAR Detection Network Based on Adaptive Scale Distribution Attention

Deep learning has made rapid progress in the field of synthetic aperture radar (SAR) detection. However, SAR images themselves have limited information, and a general detection network that is too wide and too deep can result in computational complexity and memory waste. Therefore, we design a lightweight network for multiclass SAR detection based on adaptive scale distribution attention (ASA). First, a novel backbone is designed from the perspective of lightweight model, using deep separable convolution to generate high-quality feature maps of protruding targets, and applying channel shuffle to improve training and detection efficiency. Second, a lightweight ASA is proposed, which can adaptively obtain the scattering information of multiscale targets, aggregate the position and contour features of the targets, and improve the detection accuracy of multiclass targets. Finally, anchor-free detection head is applied to improve the generalization ability and robustness of the model. Multiclass lightweight SAR detection network (MLSDNet) achieves a high mean average precision (mAP) of 92.99% on the newly released multiclass SAR target datasets (MSAR-1.0) with only 1.42 G FLOPs and 928.25 K Params. The mAP on SAR ship datasets such as SAR ship detection dataset (SSDD) and high-resolution SAR images dataset (HRSID) reached 99.1% and 94.7%, respectively. The mAP on the latest SAR aircraft dataset reached 97.7%, demonstrating its good generalization ability. Its performance has reached the state-of-the-art (SOTA).