Light-ResKAN: A Parameter-Sharing Lightweight KAN with Gram Polynomials for Efficient SAR Image Recognition
This work provides an efficient solution for edge-based SAR image recognition, which is incremental as it adapts existing KAN and ResNet concepts to a specific domain.
The paper tackles the challenge of deploying deep learning for SAR image recognition on resource-constrained edge devices by proposing Light-ResKAN, which achieves high accuracy (e.g., 99.09% on MSTAR) while significantly reducing computational costs, such as an 82.90× reduction in FLOPs compared to VGG16.
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.