ATRNet-STAR: A Large Dataset and Benchmark Towards Remote Sensing Object Recognition in the Wild (Preprint)

Yongxiang Liu*, Weijie Li, Li Liu*, Jie Zhou, Bowen Peng, Yafei Song, Xuying Xiong, Wei Yang, Tianpeng Liu, Zhen Liu, Xiang Li*

[Paper] [BibTex] [知乎] [Code]

ATRNet-STAR dataset contains 40 distinct target types, collected with the aim of replacing the outdated though widely used MSTAR dataset and making a significant contribution to the advancement of SAR ATR research.

SARATR-X: Toward Building a Foundation Model for SAR Target Recognition (TIP)

Weijie Li, Wei Yang*, Yunan Hou, Li Liu*, Yongxiang Liu*, and Xiang Li

[Paper] [BibTex] [知乎] [Code]

SARATR-X is a foundation model, which learns generalizable representations via self-supervised learning from large-scale unlabelled data and provides a corner stone for generic SAR target detection and classification.

Predicting Gradient is Better: Exploring Self-Supervised Learning for SAR ATR with a Joint-Embedding Predictive Architecture (ISPRS)

Weijie Li, Wei Yang, Tianpeng Liu, Yuenan Hou, Yuxuan Li, Zhen Liu, Yongxiang Liu*, Li Liu

[Paper] [BibTex] [知乎] [Code]

SAR-JEPA is a joint-embedding predictive architecture for SAR ATR that leverages local masked patches to predict the multi-scale SAR gradient representations of an unseen context.