Situation Recognition on imSitu
Top-1 Verb
- 2016-06-01 — CRF: Top-1 Verb 32.34
- 2016-12-03 — CRF + Aug: Top-1 Verb 34.12
- 2017-03-18 — RNN + Fusion: Top-1 Verb 35.9
- 2017-08-14 — GraphNet: Top-1 Verb 36.72
- 2019-10-01 — Kernel GraphNet: Top-1 Verb 43.27
- 2021-12-10 — SituFormer: Top-1 Verb 44.2
- 2022-03-30 — CoFormer: Top-1 Verb 44.66
- 2023-07-02 — ClipSitu: Top-1 Verb 47.23
- 2025-01-20 — Ours: Top-1 Verb 58.88
| Rank | Model | Top-1 Verb | Top-1 Verb & Value | Top-5 Verbs | Top-5 Verbs & Value | Paper | Code | Year |
|---|---|---|---|---|---|---|---|---|
| 11 | RNN + Fusion | 35.9 | 27.45 | 63.08 | 46.88 | Recurrent Models for Situation Recognition | 2017 | |
| 12 | CRF + Aug | 34.12 | 26.45 | 62.59 | 46.88 | Commonly Uncommon: Semantic Sparsity in Situation Recognition | my89/imSitu · thilinicooray/my_imsitu | 2016 |
| 13 | CRF | 32.34 | 24.64 | 58.88 | 42.76 | Situation Recognition: Visual Semantic Role Labeling for Image Understanding | my89/imSitu | 2016 |