Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces Light-ResKAN, a lightweight SAR image recognition model that replaces ResNet convolutions with KAN convolutions using Gram Polynomial activations. A parameter-sharing strategy is employed to reduce parameters and FLOPs while preserving unique channel features. Experiments on three SAR image datasets demonstrate that Light-ResKAN achieves high accuracy with significantly reduced computational cost compared to VGG16, making it suitable for edge deployment.
Achieve VGG16-level SAR image recognition accuracy with 83x fewer FLOPs and 164x fewer parameters by replacing convolutions with parameter-sharing KANs using Gram Polynomial activations.
Synthetic Aperture Radar (SAR) image recognition is vital for disaster monitoring, military reconnaissance, and ocean observation. However, large SAR image sizes hinder deep learning deployment on resource-constrained edge devices, and existing lightweight models struggle to balance high-precision feature extraction with low computational requirements. The emerging Kolmogorov-Arnold Network (KAN) enhances fitting by replacing fixed activations with learnable ones, reducing parameters and computation. Inspired by KAN, we propose Light-ResKAN to achieve a better balance between precision and efficiency. First, Light-ResKAN modifies ResNet by replacing convolutions with KAN convolutions, enabling adaptive feature extraction for SAR images. Second, we use Gram Polynomials as activations, which are well-suited for SAR data to capture complex non-linear relationships. Third, we employ a parameter-sharing strategy: each kernel shares parameters per channel, preserving unique features while reducing parameters and FLOPs. Our model achieves 99.09%, 93.01%, and 97.26% accuracy on MSTAR, FUSAR-Ship, and SAR-ACD datasets, respectively. Experiments on MSTAR resized to $1024 \times 1024$ show that compared to VGG16, our model reduces FLOPs by $82.90 \times$ and parameters by $163.78 \times$. This work establishes an efficient solution for edge SAR image recognition.