Browse › Computer Vision › Visual Question Answering (VQA) › InfographicVQA
InfographicVQA Benchmark (Visual Question Answering (VQA))
Visual Question Answering (VQA) is a task in computer vision that involves answering questions about an image. The goal of VQA is to teach machines to understand the content of an image and answer questions about it in natural language.
Image Source: visualqa.org
The archive carries no text for this table; the description above is the archive's text for the task Visual Question Answering (VQA). archive 2025-07-28
Over time archive 2025-07-28
The chart needs JavaScript; the table below carries every value.
Not inferred: ANLS. Points are placed at the row's paper date; 21 of 21 rows carry one.
Results archive 2025-07-28
Archive rows end at the archive snapshot, 2025-07-28: no result published after that date is in this table. Rank is the archive's row order at that snapshot; not re-ranked here. Metric values are the archive's strings. Column headers sort the table in your browser; each row keeps its archive rank.
| Paper | Code | Ran Syntology | Report | |||||
|---|---|---|---|---|---|---|---|---|
| 1 | Gemini Ultra (pixel only) | 80.3 | – | Paper | Code | 2023 | linked, not harvested | report |
| 2 | SMoLA-PaLI-X Specialist | 66.2 | ✓ | Paper | – | 2023 | no code linked | report |
| 3 | ScreenAI 5B (4.62 B params, w/ OCR) | 65.90 | ✓ | Paper | Code | 2024 | linked, not harvested | report |
| 4 | SMoLA-PaLI-X Generalist | 65.6 | ✓ | Paper | – | 2023 | no code linked | report |
| 5 | UDOP (aux) | 63.0 | ✓ | Paper | Code | 2022 | 4 of 17 ran · 13 unverified | report |
| 6 | PaLI-3 (w/ OCR) | 62.4 | – | Paper | Code | 2023 | 3 of 3 ran · 0 unverified | report |
| 7 | TILT-Large | 61.20 | ✓ | Paper | Code | 2021 | linked, not harvested | report |
| 8 | PaLI-3 | 57.8 | – | Paper | Code | 2023 | 3 of 3 ran · 0 unverified | report |
| 9 | ChatGPT 3.5 with LAPDoc Prompt (SpatialFormat) | 54.9 | – | Paper | – | 2024 | no code linked | report |
| 10 | PaLI-X (Single-task FT w/ OCR) | 54.8 | ✓ | Paper | Code | 2023 | 6 of 7 ran · 1 unverified | report |
| 11 | Claude + LATIN-Prompt | 54.51 | – | Paper | Code | 2023 | 1 of 10 ran · 9 unverified | report |
| 12 | PaLI-X (Multi-task FT) | 50.7 | ✓ | Paper | Code | 2023 | 6 of 7 ran · 1 unverified | report |
| 13 | PaLI-X (Single-task FT) | 49.2 | ✓ | Paper | Code | 2023 | 6 of 7 ran · 1 unverified | report |
| 14 | GPT-3.5 + LATIN-Prompt | 48.98 | – | Paper | Code | 2023 | 1 of 10 ran · 9 unverified | report |
| 15 | DocFormerv2-large | 48.8 | ✓ | Paper | Code | 2023 | linked, not harvested | report |
| 16 | UDOP | 47.4 | – | Paper | Code | 2022 | 4 of 17 ran · 13 unverified | report |
| 17 | DUBLIN (variable resolution) | 42.6 | ✓ | Paper | – | 2023 | no code linked | report |
| 18 | Pix2Struct-large | 40 | – | Paper | Code | 2022 | 0 of 5 ran · 5 unverified | report |
| 19 | Pix2Struct-base | 38.2 | – | Paper | Code | 2022 | 0 of 5 ran · 5 unverified | report |
| 20 | MatCha | 37.2 | – | Paper | Code | 2022 | linked, not harvested | report |
| 21 | DUBLIN | 36.82 | ✓ | Paper | – | 2023 | no code linked | report |
All 21 rows shown. 21 link to a paper page on this site; 11 are marked as using additional training data in the archive. No GitHub stars are tracked; "Code" is the first repository the archive lists for the row. The archive carries no row tags, review links or community-submitted rows for this table; none are shown. archive 2025-07-28
Syntology Ran reads "N of M ran · U unverified": of the M code samples Syntology harvested from repositories linked to that row's paper (joined by arXiv id), N executed on a synthesized input and the other U = M−N are unverified (harvested, no recorded run). It counts code from repositories linked to that row's paper, not this result: the row's number was not reproduced and nothing here is a correctness claim. The other cell texts mean no graph line for the row: "linked, not harvested" (the archive links code, Syntology has not harvested it), "no code linked" (no code link in the archive), "not matched" (the row's paper URL matched no paper on this site). 11 rows have a graph line, from 5 distinct papers; 9 rows (4 papers) have at least one sample that ran. Counting each paper once: Syntology ran 14 of 42 samples; 28 unverified. Separately, 3 of those 42 are pointer-only (licence): the site points at that code rather than redistributing it, a licence property recorded for ran and unverified samples alike; each cell's tooltip carries the row's own pointer-only count. Read from the graph 2026-09-24. Per-sample status is on the paper page.
Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections