Bongard-HOI
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Few-shot Classification | Bongard-HOI (test) | Accuracy (Unseen Act / Unseen Obj)87.21 | 12 | |
| Category-level few-shot learning | Bongard-HOI (test) | Accuracy76.41 | 5 | |
| Context-dependent Visual Reasoning | Bongard-HOI (test) | Accuracy (Seen Act, Seen Obj)66.39 | 4 |
Showing 3 of 3 rows