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

NUS-WIDE

Benchmarks

Task NameDataset NameSOTA ResultTrend
Multi-Label ClassificationNUS-WIDE (test)
mAP70.4
124
Image RetrievalNUS-WIDE (test)
MAP89.1
104
Image RetrievalNUS-WIDE
mAP85
89
Targeted Hashing AttackNUS-WIDE
t-MAP81.67
88
Multi-label RecognitionNUS-WIDE
mAP52.78
66
Image-to-Text RetrievalNUS-WIDE
mAP@5091.1
62
Text-to-Image RetrievalNUS-WIDE
mAP@5091.6
62
Image RetrievalNUS-WIDE 5000
mAP81.02
52
Image RetrievalNUS-WIDE
P@100083.91
50
Multi-label ClassificationNUS-WIDE 925/81 (unseen)
mAP (Mean Average Precision)53.2
43
Unsupervised Image HashingNUS-WIDE
mAP84.4
40
Multi-label ClassificationNUS-WIDE
mAP70.1
40
Multi-label classificationNUS-WIDE
mAP47.8
38
Cross-modal RetrievalNUS-WIDE (Average)
mAP43.4
32
Cross-modal RetrievalNUS-WIDE (Unseen)
mAP31.6
32
Cross-modal RetrievalNUS-WIDE (Seen)
mAP55.7
32
Multi-label Image ClassificationNUS-WIDE 81 concept labels (test)
F1 (Class)58.7
29
Fine-grained Image HashingNUS-WIDE (test)
Collision Probability (%)0.0025
18
Multi-label classificationNUS-WIDE 81 unseen labels (test)
mAP0.603
17
Image TaggingNUS-WIDE 81 unseen tags
mAP42.2
16
Multi-class image classificationNUS-WIDE geo-tagged (test)
mAP36.2
16
Multi-label Image ClassificationNUS-WIDE
CF1 (Top 3)57.8
15
Multi-label ClassificationNUS-WIDE
Precision @ K=330.9
14
Multi-view ClusteringNUS-WIDE
Accuracy64.84
14
Multi-view ClusteringNUS-WIDE 100% aligned
Accuracy67.25
14
Showing 25 of 52 rows