E-VQA
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Visual Question Answering | E-VQA (test) | Accuracy84.03 | 85 | |
| Lifelong Knowledge Editing | E-VQA Lifelong Sequential | Rel. Score96.67 | 72 | |
| Knowledge-Intensive Visual Question Answering | E-VQA (test) | Accuracy (All)41.4 | 34 | |
| Knowledge-Based Visual Question Answering | E-VQA Single-Hop | Accuracy59.9 | 27 | |
| Knowledge Editing | E-VQA MMEdit 1.0 (test) | Reliability100 | 24 | |
| Knowledge-based VQA | E-VQA | Single-Hop Accuracy70.27 | 16 | |
| Knowledge-Based Visual Question Answering | E-VQA All | Accuracy45.8 | 15 | |
| Visual Question Answering | E-VQA | Accuracy53.7 | 15 | |
| Conflict Discrimination | E-VQA (test) | MCC93.4 | 12 | |
| Model Editing | E-VQA 5 | Reliability Score98.12 | 11 | |
| Retrieval | E-VQA | Recall@144.9 | 11 | |
| Lifelong Editing | E-VQA Lifelong Editing 5 | Relational Score96.85 | 10 | |
| Re-ranking | E-VQA | CondR@190 | 10 | |
| Re-ranking | E-VQA | Recall@144.7 | 10 | |
| Retrieval | E-VQA standard (test) | R@144.1 | 8 | |
| Entity Retrieval | E-VQA (test) | Recall@10.428 | 7 | |
| Knowledge Editing | E-VQA | Reliability98.12 | 6 | |
| Open-domain visual recognition | E-VQA (Overall) | Top-1 Accuracy47.4 | 6 | |
| Knowledge-based Visual Retrieval | E-VQA 1.0 (test) | MRR@50.4492 | 6 | |
| Knowledge Editing | E-VQA 1,000 sequential edits | Reliability96.85 | 5 | |
| Knowledge Retrieval | E-VQA (test) | R@2058.7 | 5 | |
| Retrieval | E-VQA (test) | PR@50.781 | 5 | |
| Multimodal Multi-hop Visual Question Answering | E-VQA Two hop | Accuracy23.3 | 4 | |
| Retrieval | E-VQA M2KR | R@143.1 | 3 | |
| Open-domain visual recognition | E-VQA (Seen) | Top-1 Accuracy39.9 | 3 |
Showing 25 of 28 rows