Visual Question Answering on WebQA (val)
53FLPROMPTCAP + GPT-3
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| PROMPTCAP + GPT-3Evaluation protocol=8-shot in-context learning2022.11 | 53 | 57.2 | 34.5 | |
| OFA-Cap + GPT-3Evaluation protocol=8-shot in-context learning2022.11 | 52.8 | 55.4 | 33.5 | |
| VLP + VinVLEvaluation protocol=Fully supervised2022.11 | 47.6 | 49.6 | 27.5 | |
| VLP + x101fpnEvaluation protocol=Fully supervised2022.11 | 46.9 | 44.3 | 23.8 |