A self-supervised Vision Transformer trained on SETHI radar patches, without any MSTAR labels, gives 95.9% few-shot classification accuracy on MSTAR using k-NN with ten labels per class.
It underscores great adaptability across sensor types
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
eess.SP 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
General Feature Extraction In SAR Target Classification: A Contrastive Learning Approach Across Sensor Types
A self-supervised Vision Transformer trained on SETHI radar patches, without any MSTAR labels, gives 95.9% few-shot classification accuracy on MSTAR using k-NN with ten labels per class.