Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Cross-modal Retrieval on InstVL video (Global)
Loading...
94.5
T2V R@1
InstAP
33.2336
49.1393
65.045
80.9507
Apr 9, 2026
T2V R@1
V2T R@1
Updated 9d ago
Evaluation Results
Method
Method
Links
T2V R@1
V2T R@1
InstAP
2026.04
94.5
95.5
UMT-L
2026.04
88.3
85.5
UMT-L (InstVL; g)
training_corpus=InstVL...
2026.04
84.8
82.4
VideoPrism
2026.04
82.71
83.62
OpenCLIP
2026.04
82
77.15
UMT-L (InstVL; g+i)
training_corpus=all In...
2026.04
79.9
77.2
SigLIP
2026.04
74.72
76.14
CLIP4Clip
2026.04
67.5
70.5
ViCLIP
2026.04
62.89
62.69
MCQ
2026.04
61.48
60.67
Coca
2026.04
46.92
43.78
CLIP-ViP
2026.04
35.59
61.07
Feedback
Search any
task
Search any
task