Benchmark
FG-OVD
A fine-grained open-vocabulary detection benchmark dataset focused on distinguishing fine-grained subcategories within the same broad category. It uses average precision (AP) as the evaluation metric and requires models to accurately localize and classify unseen fine-grained categories.
Timeline (last 90 days)
Oct 20
On the FG-OVD benchmark, SL-Ground achieves the best average AP among compared methods.
Unverified50%