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

Flower

Benchmarks

Task NameDataset NameSOTA ResultTrend
Fine-grained classificationFlower102 (test)
Accuracy97.8
27
Image ClassificationFlower102
Base Accuracy96.5
26
Fine-grained classificationFlower102
Clean Accuracy76.5
25
Image ClassificationFlower
Accuracy95.84
23
Fine-grained Image ClassificationFlower102
Accuracy99.19
19
Image ClassificationFlower (test)
Accuracy99.45
18
Image ClassificationFlower102
Accuracy99
16
Image ClassificationFlower-102 (test)
NLL (N=510)0.54
16
Image ClassificationFlower102
Clean Accuracy65.6
13
Image ClassificationFlower102
Robust Accuracy37.02
13
Fine-grained ClassificationFlower102
Top-1 Acc98.9
13
Spectral ClusteringFlower-20M
Time Cost (s)311.94
11
ClusteringFlower-20M
Cluster Accuracy93.79
11
Multi-view ClassificationFlower (test)
Balanced Accuracy0.8633
10
Fine-grained Visual RecognitionFlower-102 (test)
EM96.5
9
Equilibrated reaction predictionFlowER 19 (original)
Step Accuracy (Top-1)89.74
8
Long-tail classificationFlower-LT IF=10
Accuracy97.8
7
Long-tail classificationFlower-LT IF=50
Accuracy96.85
7
Long-tail classificationFlower-LT IF=100
Accuracy (Many)100
7
Image GenerationFlower
FID11.5
7
Ensemble ClusteringFlower-20M
Time cost (s)764.09
7
Ensemble ClusteringFlower-20M
NMI (%)92.47
7
Image ClassificationFlower
Accuracy97.7
6
Fine-grained Visual RecognitionFlower-102
cACC67.2
5
Image SegmentationFlower
Accuracy90.1
5
Showing 25 of 34 rows