Tiny-ImageNet
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Tiny-ImageNet 1.0 (test)
67.64Accuracy (Natural)
75
Tiny-ImageNet (val)
0.113Median L2 Distance
64
Tiny-ImageNet 20 tasks
67.95Average Accuracy
54
Tiny ImageNet 200 (test)
66.16Test Accuracy
53
Tiny-ImageNet 1 class (forget)
66.52Retain Accuracy
48
Tiny-ImageNet-C 30% corrupted (test)
43.7Accuracy
45
Seq-Tiny-ImageNet
69.77Final Average Accuracy
44
Tiny-ImageNet fine-grained
21.36Accuracy
42
Tiny-ImageNet fine-grained (test)
22.95Accuracy
42
Tiny-ImageNet (test)
0.272Relative Forgetting
36
Split Tiny-ImageNet (test)
82.04Average Accuracy (A_T)
35
Tiny-ImageNet Dirichlet-0.05 (test)
54Accuracy
32
Tiny-ImageNet 10 tasks
49.4Accuracy
31
Tiny ImageNet 64x64 (test)
29.6Accuracy
27
Tiny-ImageNet (train)
99.52Precision
24
Tiny-ImageNet heterogeneous (test)
54.25Accuracy
24
S-Tiny-ImageNet (last)
82.04TIL Last Top-1 Acc
24
S-Tiny-ImageNet Task-IL
82.04Accuracy
24
Tiny-ImageNet corrupted (test)
30.41Accuracy
23
S-Tiny-ImageNet Class-IL
59.99Accuracy
23
Tiny ImageNet Near-OOD (test)
26.53FPR@95
22
Tiny-ImageNet Split (test)
59.99Average Accuracy (AT)
22
Tiny-ImageNet 20 classes (forget)
66.08Accuracy
21
Tiny-ImageNet 1 class forget (Retain set)
66.52Accuracy
21
Tiny-ImageNet LT (test)
94.67Accuracy (Dogs)
20