Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Visual Grounding on ScanRefer (test, Overall Category)
Loading...
61.1
Accuracy
DINOv3 + SpatialBoost
49.66
52.63
55.6
58.57
Mar 23, 2026
Accuracy
Updated 25d ago
Evaluation Results
Method
Method
Links
Accuracy
DINOv3 + SpatialBoost
SpatialBoost=true
2026.03
61.1
DINOv2 + SpatialBoost
SpatialBoost=true
2026.03
57
SigLIPv2 + SpatialBoost
SpatialBoost=true
2026.03
56.8
OpenCLIP + SpatialBoost
SpatialBoost=true
2026.03
56.6
DINOv3
2026.03
56.2
V-JEPAv2
Encoder Type=Vision-only
2026.03
55.5
DINOv2
2026.03
52.7
dino.txt
Encoder Type=Vision-La...
2026.03
52.5
TIPS
Encoder Type=Vision-La...
2026.03
52
PE-Core
Encoder Type=Vision-La...
2026.03
51.7
SigLIPv2
2026.03
51.4
AIMv2
Encoder Type=Vision-La...
2026.03
50.9
OpenCLIP
2026.03
50.1
Feedback
Search any
task
Search any
task