Fashion MNIST
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Fashion-MNIST
94.53Top-1 Accuracy
30
Fashion MNIST (test)
-233.39Log-Likelihood
30
Fashion MNIST (train)
-230.77Log Likelihood (100 samples)
30
Fashion-MNIST Sample Unlearning (Df)
100Accuracy
28
Fashion MNIST (test)
1.005Accuracy Variance
28
Fashion-MNIST (test)
0.922SC
28
Fashion-MNIST Dir(0.5) (test)
90.49Accuracy
28
Fashion-MNIST under sudden drift and Dir(0.1) 1.0 (test)
87.63Accuracy (Generalized)
28
Fashion-MNIST Dir(0.5)
90.57Accuracy
28
Fashion-MNIST
6Avg Malicious Clients Detected
24
Fashion-MNIST
97SSIM
24
Rotated Fashion MNIST (target domains 0° and 90°)
81.6Accuracy
21
Fashion-MNIST
93.89Accuracy
20
Fashion-MNIST Non-IID Dir(0.3) 2017 (test)
89.31Top-1 Accuracy
20
Fashion-MNIST Non-IID Dir(0.5) 2017 (test)
90.79Top-1 Accuracy
20
Fashion-MNIST IID 2017 (test)
92.43Top-1 Accuracy
20
Fashion-MNIST
65.93Pre-training Accuracy
20
Fashion-MNIST (test)
99.2Pre-training Accuracy
20
Fashion-MNIST (test)
82.4Accuracy
20
Fashion-MNIST heterogeneous (test)
83Test Accuracy
18
Fashion-MNIST standard (test)
96.3Accuracy
18
Fashion-MNIST
732Execution Time (s)
18
Fashion-MNIST standard (test)
65.8ARI
17
Fashion-MNIST
2.2Activation Distance (Dr)
16
Fashion-MNIST
0Activation Distance (Dr)
16