Our Bench
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Our-Bench SpatialTree-Bench 1.0 (test)
57.8Average Score
16
Our Bench
76.34Accuracy
8
Our Bench
32.33CLIP Score
6
Our Bench Layout only
44F1 Score
5
Our Bench Layout + Reference
35F1 Score
4