Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

MSTAR

Benchmarks

Task NameDataset NameSOTA ResultTrend
Federated Learning Image ClassificationMSTAR
Accuracy95.13
99
Backdoor Attack RobustnessMSTAR (test)
ASR99.37
90
SAR Image ClassificationMSTAR (test)
ACC99.09
66
Adversarial AttackMSTAR DARPA (test)
ASR85.61
42
ClassificationMSTAR 8-shot
Precision72.5
25
ClassificationMSTAR 4-shot
Precision60.65
25
ClassificationMSTAR 2-shot
Precision52.98
25
SAR ClassificationMSTAR SOCs (Standard operating conditions) 10-way
Accuracy95.9
25
SAR Automatic Target RecognitionMSTAR
Accuracy99.59
13
SAR Image RecognitionMSTAR N-Shot 20-shot (test)
Accuracy59.51
12
SAR Image RecognitionMSTAR 10-shot N-Shot (test)
Accuracy52.57
12
SAR Image RecognitionMSTAR N-Shot 5-shot (test)
Accuracy41.57
12
SAR Image ClassificationMSTAR resized to 1024 x 1024 1.0 (test)
Accuracy99.46
12
SAR ClassificationMSTAR EOCs (Target version variations) 4-way
Accuracy82.8
12
SAR ClassificationMSTAR EOCs (Target configuration variations) 4-way
Accuracy78.3
12
SAR ClassificationMSTAR EOCs (Depression angle variations) 4-way
Accuracy98.9
12
Visual Question AnsweringMSTAR
Accuracy40.9
11
ClassificationMSTAR
Macro-F191.53
9
Adversarial AttackMSTAR (test)
PSNR (dB)31.48
7
Target ClassificationMSTAR 30% label
Top-1 Accuracy99.61
6
Target ClassificationMSTAR 40-shot
Top-1 Accuracy97.24
5
Image ClassificationMSTAR 100%
Accuracy0.961
5
Image ClassificationMSTAR (90%)
Accuracy95.6
5
Image ClassificationMSTAR (50%)
Accuracy92.3
5
Image ClassificationMSTAR (5%)
Accuracy69.5
5
Showing 25 of 25 rows