View Intervention and Feature Alignment Aggregation Framework for Multiview SAR Target Recognition

Multiview synthetic aperture radar (SAR) automatic target recognition (ATR) has attracted increasing attention for its ability to integrate effective information from multiple images. However, the existing algorithms have ignored the interplay between the multiview combination and the multiview network, failing to explore the inherent coupling relationship within multiview images. To tackle these issues, a multiview SAR ATR framework called view intervention and feature alignment aggregation is proposed. First, a deep clustering-based multiview combination is designed. Images with sufficient complementary information are selected from the raw SAR data under each category to form multiview images according to image features, which are the latent features obtained by the autoencoder (AE). Next, an efficient multiview feature alignment aggregation (Mv-FAA) network is proposed, in which the encoder of the AE serves as the feature extraction module. By designing a hybrid loss function to guide the training of the Mv-FAA network, it can extract complementary features from multiview images while retaining certain consistent features so that the final holistic features of the target are obtained for discrimination. The proposed framework strengthens the link between the multiview combination and the multiview network to reconcile the complementary and consistent information within multiview images, providing valuable insights for advancing multiview SAR ATR research. The experimental results on the Moving and Stationary Target Recognition and the Full Aspect Stationary Targets-Vehicle datasets have achieved state-of-the-art performance.