Browse › Computer Vision › Visual Question Answering (VQA) › GQA Test2019

Visual Question Answering (VQA) archive 2025-07-28

GQA Test2019 Benchmark (Visual Question Answering (VQA))

127 rows 5 with code listed 7 metrics Dataset page

Syntology Since the archive (Syntology): none placed; 5 newer papers judged to report on this board so far.

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

Direction inferred from the metric name, not from the archive: Accuracy (higher is better). Not inferred (points only, no best-so-far line): Binary, Open, Consistency, Plausibility, Validity, Distribution. Points are placed at the row's paper date, not an evaluation date; 6 of 127 rows carry one.

The chart needs JavaScript; the table below carries every value.

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 human 89.391.287.498.497.298.90.0 – – – not matched report
2 DREAM+Unicoder-VL (MSRA) 76.0484.4668.691.4783.7596.423.68 – – – not matched report
3 TRRNet (Ensemble) 74.0382.1266.8989.083.5896.761.29 – – – not matched report
4 MIL-nbgao 73.8180.867.6491.7683.996.731.7 – – – not matched report
5 Kakao Brain 73.3379.6867.7377.0283.796.362.46 – – – not matched report
6 Coarse-to-Fine Reasoning, Single Model 72.1481.1664.1990.9684.8196.772.39 – – – not matched report
7 270 70.2377.563.8286.9483.7796.651.49 – – – not matched report
8 NSM ensemble (updated) 67.5580.4556.1693.8384.1696.532.78 – – – not matched report
9 VinVL-DPT 64.9282.6349.2994.3784.9196.645.11 – – – not matched report
10 VinVL+L 64.8582.5949.1994.084.9196.624.59 – Paper Code 2023 linked, not harvested report
11 Single Model 64.6582.6348.7794.3584.9896.624.72 – Paper Code 2021 community repositories only · 2 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 1 where Syntology's instrument failed) · 0 unverified report
12 Wayne 63.9480.8449.0391.5484.7496.564.69 – – – not matched report
13 Single 63.277.9150.2289.8485.1596.475.25 – – – not matched report
14 NSM single (updated) 63.1778.9449.2593.2584.2896.413.71 – – – not matched report
15 LXR955, Ensemble 62.7179.7947.6493.185.2196.366.42 – Paper Code 2019 community repositories only · 4 ran (of which 4 constructed an object rather than computing a result; 4 with no instrument failure: 0 honoured, 0 violated, 4 with no contract checked; 0 where Syntology's instrument failed) · 11 unverified; every one of the 4 samples that ran constructed an object rather than computing a result report
16 MDETR 62.4580.9146.1593.9584.1596.335.36 – – – not matched report
17 1-gqa 62.4480.2846.6994.3684.9196.465.33 – – – not matched report
18 UCM 61.4978.446.5688.6884.8596.335.7 – – – not matched report
19 GRN 61.2278.6945.8190.3185.4396.366.77 – Paper – 2019 no code linked report
20 lxmert-adv-txt 61.1278.0746.1691.1384.896.365.55 – – – not matched report
21 lxmert-adv-txt 61.177.9946.1991.0884.8296.365.52 – – – not matched report
22 MSM@MSRA 61.0977.8446.388.9285.4996.435.68 – – – not matched report
23 mlmbert 61.0578.0246.0689.7784.9596.55.24 – – – not matched report
24 fisher 60.9877.3246.5590.7784.9396.385.36 – – – not matched report
25 ckpt 19 exp 90 60.9578.4145.5489.0884.2796.354.86 – – – not matched report
26 45 60.9377.8346.0190.384.6996.355.74 – – – not matched report
27 IQA (single) 60.8978.0745.7393.0284.0596.05.31 – – – not matched report
28 Ensemble10 60.8779.1244.7692.6185.6396.358.56 – – – not matched report
29 Meta Module, Single 60.8378.944.8992.4984.5596.195.54 – – – not matched report
30 xpj 60.777.4145.9689.6584.5596.376.09 – – – not matched report
31 fbe20v3.json 60.6778.0245.3689.8184.8496.316.41 – – – not matched report
32 LININ 60.5978.4444.8392.6685.3896.577.28 – – – not matched report
33 prompt IMT-16 60.5176.8746.0688.285.1996.158.48 – – – not matched report
34 vv69 60.4277.1245.6889.6984.5696.356.03 – – – not matched report
35 bert_v1 60.3777.0945.6189.7784.5696.226.43 – – – not matched report
36 LXR955, Single Model 60.3377.1645.4789.5984.5396.355.69 – Paper Code 2019 community repositories only · 4 ran (of which 4 constructed an object rather than computing a result; 4 with no instrument failure: 0 honoured, 0 violated, 4 with no contract checked; 0 where Syntology's instrument failed) · 11 unverified; every one of the 4 samples that ran constructed an object rather than computing a result report
37 IIE_Morningstar 60.2877.1345.4189.4784.4596.335.38 – – – not matched report
38 full_nsp_ft_results_submit_predict.json 60.2776.9945.5190.1684.4996.315.39 – – – not matched report
39 TESTOVQA007 60.1876.9745.3689.6584.4796.335.29 – – – not matched report
40 test gqa 60.1876.8445.4889.7784.696.375.65 – – – not matched report
41 Future_Test_team 60.1777.1945.1489.6184.4696.365.83 – – – not matched report
42 tmp 60.1477.1545.1289.5884.4796.365.81 – – – not matched report
43 Inspur 60.0776.8445.2789.3284.5596.356.21 – – – not matched report
44 full_nsp_mlm_ft_joint_results_submit_predict.json 60.0276.3745.5990.0584.3496.295.63 – – – not matched report
45 SSRP 60.0176.7745.2189.1784.4696.356.28 – – – not matched report
46 Musan 59.9379.0943.0293.7285.9296.4110.1 – – – not matched report
47 gaochongyang9 59.8476.7944.8989.5284.7296.26.06 – – – not matched report
48 PVR 59.8178.0243.7591.4384.7796.56.0 – – – not matched report
49 BgTest 59.876.7444.8589.1484.296.235.11 – – – not matched report
50 DAM 59.7277.9743.6189.4384.8996.556.25 – – – not matched report
51 DL16 59.5477.9843.2689.2184.9496.246.01 – – – not matched report
52 mcmi 59.4377.1143.8289.0584.9496.566.39 – – – not matched report
53 rishabh_test 59.3777.5343.3588.6384.7196.186.06 – – – not matched report
54 UNITER + MAC + Graph Networks 59.2977.3143.3888.9484.4396.35.8 – – – not matched report
55 LXMERT-S 59.1276.6943.688.984.7896.435.6 – – – not matched report
56 QGCRGN 59.0676.0744.0489.8182.7693.826.14 – – – not matched report
57 gbert1 58.9176.0843.7589.5284.5296.186.93 – – – not matched report
58 glimple_all 58.8875.0744.5884.6484.8696.235.54 – – – not matched report
59 ours-4-gqa_el_tag_v4__pretrain_rel_tag_dist_tc_v7_checkpoint-47-157510-best-4.json 58.7276.443.1189.5884.6896.216.58 – – – not matched report
60 Partial-MSP 58.4277.3941.6790.2984.5395.577.86 – – – not matched report
61 UCAS-SARI 58.275.9142.5788.2584.7296.085.81 – – – not matched report
62 stu09e 58.1276.3942.088.0184.896.065.65 – – – not matched report
63 happyTeam 58.0676.641.790.9685.2796.317.6 – – – not matched report
64 graphRepresentation, Single 57.8974.5443.1985.4584.9996.45.73 – – – not matched report
65 VqaStar-UCAS-SARI 57.7975.3742.2688.384.8596.115.65 – – – not matched report
66 REX 57.7775.7841.8686.8584.9796.445.36 – – – not matched report
67 MLVQA (single) 57.6575.2242.1487.3584.7396.185.48 – – – not matched report
68 rsa-14word 57.3575.0741.7187.6184.595.865.94 – – – not matched report
69 result_run_2647872_epoch11 57.2174.4641.9987.684.8796.25.6 – – – not matched report
70 DeeTee 57.1475.0741.3187.3684.4995.875.29 – – – not matched report
71 BAN 57.176.040.4191.785.5896.1610.52 – – – not matched report
72 LCGN 57.0773.7742.3384.6884.8196.484.7 – – – not matched report
73 RSN (Single Model) 57.0174.7841.3287.7484.2596.036.06 – – – not matched report
74 GM6_9_2_train 56.9674.9741.0685.1284.8596.387.13 – – – not matched report
75 wcf-fight 56.9575.0141.0290.4985.4696.379.5 – – – not matched report
76 total14 56.9574.6241.3687.7184.5795.985.81 – – – not matched report
77 Testify 56.6573.6541.6484.3584.3795.946.07 – – – not matched report
78 F205 56.5973.042.1184.784.8696.44.68 – – – not matched report
79 Feb_ft2_mergeadd_weightalllstm_picklocw_box5_prep 56.3874.8440.0991.7183.7695.436.32 – – – not matched report
80 MMT-VQA 56.2873.7340.8786.8684.296.015.78 – – – not matched report
81 IWantADonut 56.1872.8441.4785.4684.0496.185.42 – – – not matched report
82 GIN 56.1673.5640.884.9984.8396.45.87 – – – not matched report
83 LOGNet+VLR 56.1172.6541.5285.5184.3696.255.42 – – – not matched report
84 Improved SNMN 56.0973.440.8285.1184.7996.375.14 – – – not matched report
85 ST_VQA 56.073.940.287.1684.4596.016.02 – – – not matched report
86 RD 55.9371.8141.9383.285.0996.016.05 – – – not matched report
87 Deepblue_Semantics 55.772.8840.5383.5284.8196.395.32 – – – not matched report
88 LW 55.6572.8640.4689.1885.2796.339.69 – – – not matched report
89 RSN (Single Model)_v6 55.5772.3940.7483.3284.2496.1510.18 – – – not matched report
90 nogg 55.4172.8739.9983.0684.7496.355.48 – – – not matched report
91 abc_test 55.3572.6540.0884.1784.5696.325.22 – – – not matched report
92 KU 55.072.0939.9283.4784.6696.345.29 – – – not matched report
93 Eden_test 54.9471.740.1482.7184.7896.45.1 – – – not matched report
94 HDU_ZWF 54.7972.4239.2386.184.5595.926.01 – – – not matched report
95 vips 54.1569.340.7982.3685.1595.995.41 – – – not matched report
96 MAC 54.0671.2338.9181.5984.4896.165.34 – – – not matched report
97 5TMT-qe+o 53.8972.5237.4487.4785.0596.398.66 – – – not matched report
98 ZhaoLab 53.8568.4440.9780.285.1996.285.84 – – – not matched report
99 test 53.5770.1538.9481.1484.6796.365.32 – – – not matched report
100 Sorbonne 53.3170.4138.2380.3384.3295.996.4 – – – not matched report
101 UJCNN 52.368.4638.0484.3685.296.212.54 – – – not matched report
102 MJ 52.1969.1537.2278.3483.4495.455.69 – – – not matched report
103 mac_qin 52.0267.3538.580.4483.9495.755.64 – – – not matched report
104 Mithrandir 51.8767.9937.6480.284.3596.256.77 – – – not matched report
105 happy 51.5167.8237.1179.783.6995.826.16 – – – not matched report
106 Space Cat 51.2269.3635.282.4483.8296.126.45 – – – not matched report
107 BottomUp 49.7466.6434.8378.7184.5796.185.98 – Paper Code 2017 official: no sample here; runs from other or unrecorded repositories · 9 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 1 violated, 0 with no contract checked; 8 where Syntology's instrument failed) · 0 unverified report
108 RAM_BUGGY 49.2867.5933.1283.6883.4194.9514.28 – – – not matched report
109 sparsemax15 49.2766.5734.078.5184.5895.786.91 – – – not matched report
110 mfb+bert 48.9763.8535.8383.8583.9395.6213.72 – – – not matched report
111 RES 48.4465.0233.8181.1985.2996.1517.79 – – – not matched report
112 LAS 47.7266.2831.3484.1684.5295.4519.05 – – – not matched report
113 test 47.3858.7637.3473.7181.7594.556.29 – – – not matched report
114 LSTM+CNN 46.5563.2631.874.5784.2596.027.46 – – – not matched report
115 113 45.8664.7429.270.5786.1396.618.38 – – – not matched report
116 Ediburgh-Mila-UCLA 44.0657.5732.1338.1875.1985.948.35 – – – not matched report
117 bear 43.8459.2430.2467.7184.0195.3210.99 – – – not matched report
118 CHAIR 42.7561.2126.4563.5184.295.997.63 – – – not matched report
119 MReaL 41.6355.1229.7382.2177.492.2713.01 – – – not matched report
120 LSTM 41.0761.922.6968.6887.396.3917.93 – – – not matched report
121 Academia Sinica 40.361.1821.8874.1186.1396.1440.44 – – – not matched report
122 Fj 37.0356.6119.7463.9685.1295.7628.4 – – – not matched report
123 Mycsulb 36.7555.2420.4469.9384.1395.140.84 – – – not matched report
124 LocalPrior 31.2447.916.6654.0484.3184.3313.98 – – – not matched report
125 GlobalPrior 28.942.9416.6251.6974.8188.8693.08 – – – not matched report
126 muc_ai 26.4545.699.4755.2350.9360.8111.49 – – – not matched report
127 CNN 17.8236.051.7462.434.8435.7819.99 – – – not matched report

All 127 rows shown. 6 link to a paper page on this site; 0 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 ran (of which C constructed an object rather than computing a result; K with no instrument failure: H honoured, V violated, P with no contract checked; I where Syntology's instrument failed) · U unverified”: of the code samples Syntology harvested from repositories linked to that row's paper (joined by arXiv id), N executed on a synthesized input and U are unverified (harvested, no recorded run). Of the N, C constructed an object rather than computing a result (a part of N, never taken away from it), K ran with no instrument failure (honoured or violated the contract Syntology drafted, or ran with no contract checked) and I are failures of Syntology's instrument, not of the code (a wrong contract draft, or a contract check that could not 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). 4 rows have a graph line, from 3 distinct papers. Rows whose paper has a sample that ran: 4 (4 with a run with no instrument failure, 0 where every run was a failure of Syntology's instrument). Papers: 3 with a sample that ran: 3 with a run with no instrument failure, 0 where every run was a failure of Syntology's instrument. Whose code ran. When the archive marks a repository official for the row's paper, the cell starts with that repository's state (“official” is the archive's flag, not a verdict on who wrote the code): “official (archive's flag): N ran” counts the samples from it that ran, instrument failures included; “official: harvested, nothing ran”; when it has no sample for this paper and other samples ran, “community repositories only” if every one came from a repository the archive lists for the paper but neither marks official nor records as mentioned in it, or one Syntology harvested outside the archive's links, and “official: no sample here; runs from other or unrecorded repositories” if any came from a repository named in the paper or found in its text by Syntology, or has no repository recorded; “official: harvested for another paper” (no sample from it is linked to this paper) or “official: not harvested”. No state is printed when the archive marks none official. The cell's tooltip names the repositories the samples that ran came from. Of the 4 rows with a graph line, 4 have an archive-official repository: on 0 a sample from it ran (0 with a run with no instrument failure), on 3 the samples that ran came from community repositories only, and on 1 from other or unrecorded repositories. The filters above hide rows and never renumber them; the first two count, by default, only a run with no instrument failure, and offer any run, instrument failures included. Counting each paper once, of 26 samples: 15 ran (of which 4 constructed an object rather than computing a result; 6 with no instrument failure: 0 honoured, 1 violated, 5 with no contract checked; 9 where Syntology's instrument failed) · 11 unverified. Separately, 10 of those 26 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. Syntology's record for this page last changed ; when this build read Syntology's graph is in the build record. Per-sample status is on the paper page; each Ran cell links to that paper's sample list. For agents: get_harvested_code_for_paper(arxiv_id) lists each paper's samples; how to connect.

Since the archive Syntology

Of the 8,888 newer papers whose tables Syntology has read (8,886 checked in full and 2 with a leaderboard still awaiting a verdict), 52 had this board among the leaderboards the extractor considered for their tables (it keeps at most 10 boards per paper and 4 per dataset; the lists of 1,558 of the 8,888 were already full without this board, so if it was crowded out of them it is not counted), and 5 proposed a result; of those 5, 5 gave no cell that passed the extractor's own checks, 0 were rejected by the independent check, 0 await a verdict and 0 were placed. Syntology has checked 8,886 of the 18,637 papers on this site that are newer than the archive (for 3,717 of them no archive leaderboard matched the paper's tables, so there was nothing further to check); 733 were read and have nothing to place (arXiv has no HTML version of the paper, or that version has no tables), 41 could not be read (the extractor's reply could not be parsed), and results from the other 8,977 appear after they are checked.

Newer papers whose tables Claude Sonnet 4.5 judged to report on this benchmark (a model's judgement, not a result; newest first by arXiv id)

Paper arXiv date Ran Syntology Outcome
What Transfers from Text to Vision? Capability Scaling Laws and Transfer Dynamics for VLMs arXiv:2608.00013 24 Jun 2026 repository linked, no samples harvested no cell passed the extractor's checks
CLIP Tricks You: Training-free Token Pruning for Efficient Pixel Grounding in Large... arXiv:2605.13178 13 May 2026 15 harvested: 11 ran (of which 0 constructed an object rather than computing a result; 6 with no instrument failure: 3 honoured, 0 violated, 3 with no contract checked; 5 where Syntology's instrument failed) · 4 unverified refused by a rule: one configuration gave two values for one column
Generative Semantic Multi-Object Tracking: A Large-Scale Benchmark and an MLLM-Driven... arXiv:2601.06550 10 Jan 2026 2 harvested: 2 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 2 where Syntology's instrument failed) · 0 unverified refused by a rule: the row is an ablation variant, not the method itself
CoIDO: Efficient Data Selection for Visual Instruction Tuning via Coupled... arXiv:2510.17847 11 Oct 2025 14 harvested: 8 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 1 honoured, 0 violated, 2 with no contract checked; 5 where Syntology's instrument failed) · 6 unverified refused by a rule: the table's column header does not name the leaderboard's metric
Discrete Guidance Matching: Exact Guidance for Discrete Flow Matching arXiv:2509.21912 26 Sep 2025 6 harvested: 6 ran (of which 0 constructed an object rather than computing a result; 5 with no instrument failure: 0 honoured, 0 violated, 5 with no contract checked; 1 where Syntology's instrument failed) · 0 unverified refused by a rule: the table's column header does not name the leaderboard's metric

Syntology 5 newer papers judged to report on this board: the papers counted above as having proposed a result (0 placed); every one is listed here and in the JSON twin. The order is the arXiv identifier's, newest first, not a rank. The judgement is the model's, read from the paper's own tables; the outcome is what Syntology's checks did with it, and no value from a paper that was not placed is printed. “refused by a rule” names the extractor's rule; “no cell passed the extractor's checks” means the extractor's record holds no cell for this board, passed or refused. The 5 counted above as giving no cell that passed the extractor's own checks are the 4 marked “refused by a rule” and the 1 marked “no cell passed the extractor's checks” together. Drawn from 8,886 of 18,637 newer papers checked so far (733 more had nothing to place; 41 could not be read).

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