Knowledge-Based Visual Question Answering (Direct Answer) on A-OK-VQA (val)
0.586AccuracyOurs + MCAN (LLaMA 2)
Evaluation Results
| Method | Links | |
|---|---|---|
| Ours + MCAN (LLaMA 2)Backbone=LLaMA 2, Shot Selection Strategy=MCAN, Number of Captions=9, Query Ensembles=52023.10 | 0.586 | |
| ProphetShot Selection Strategy=MCAN2023.10 | 0.582 | |
| Ours + MCAN (LLaMA)Backbone=LLaMA, Shot Selection Strategy=MCAN2023.10 | 0.574 | |
| Ours (LLaMA 2)Backbone=LLaMA 22023.10 | 0.571 | |
| Ours (LLaMA)Backbone=LLaMA2023.10 | 0.544 | |
| GPV-22023.10 | 0.486 | |
| KRISP2023.10 | 0.337 | |
| ClipCap2023.10 | 0.309 | |
| LXMERT2023.10 | 0.307 | |
| ViLBERT2023.10 | 0.306 |