Gamma Distribution PCA-Enhanced Feature Learning for Angle-Robust SAR Target Recognition

Scattering characteristics of synthetic aperture radar (SAR) targets are typically related to observed azimuth and depression angles. However, in practice, it is difficult to obtain adequate training samples at all observation angles, which probably leads to poor robustness of deep networks. In this paper, we first propose a Gamma-Distribution Principal Component Analysis ($\Gamma$PCA) model that fully accounts for the statistical characteristics of SAR data. The $\Gamma$PCA derives consistent convolution kernels to effectively capture the angle-invariant features of the same target at various attitude angles, thus alleviating deep models' sensitivity to angle changes in SAR target recognition task. We validate $\Gamma$PCA model based on two commonly used backbones, ResNet and ViT, and conduct multiple robustness experiments on the MSTAR benchmark dataset. The experimental results demonstrate that $\Gamma$PCA effectively enables the model to withstand substantial distributional discrepancy caused by angle changes. Additionally, $\Gamma$PCA convolution kernel is designed to require no parameter updates, introducing no extra computational burden to the network. The source code is available at \href{https://github.com/ChGrey/GammaPCA}{https://github.com/ChGrey/GammaPCA}.

In the context of radar target recognition, target structures are typically related to observing angles of radar towards target. Consequently, if there are not adequate training samples at all observation angles, the recognition performance of deep networks is generally unsatisfying. To tackle this problem, we propose a Gamma-PCA model to extract intrinsic features of radar target, thereby alleviating the sensitivity of deep networks to variations in imaging angles. Think of it like recognizing a friend whether they're facing you or turned sideways — our method enables the deep networks this capability in the recognition of radar targets. The proposed Gamma-PCA can be integrated with commonly used backbones, including ResNet and ViT. Experiments on a standard radar dataset demonstrate that our method significantly improves recognition reliability, even under large variations in viewing angles. Moreover, Gamma-PCA introduces no additional training overhead, ensuring computational efficiency. The source code is available at https://github.com/ChGrey/GammaPCA.