HyperAI

Visual Question Answering On Gqa Test2019

Metrics

Accuracy
Binary
Consistency
Distribution
Open
Plausibility
Validity

Results

Performance results of various models on this benchmark

Model Name
Accuracy
Binary
Consistency
Distribution
Open
Plausibility
Validity
Paper TitleRepository
fisher60.9877.3290.775.3646.5584.9396.38--
LININ60.5978.4492.667.2844.8385.3896.57--
LSTM+CNN46.5563.2674.577.4631.884.2596.02--
VinVL-DPT64.9282.6394.375.1149.2984.9196.64--
rishabh_test59.3777.5388.636.0643.3584.7196.18--
BgTest59.876.7489.145.1144.8584.296.23--
Future_Test_team60.1777.1989.615.8345.1484.4696.36--
UNITER + MAC + Graph Networks59.2977.3188.945.843.3884.4396.3--
Fj37.0356.6163.9628.419.7485.1295.76--
rsa-14word57.3575.0787.615.9441.7184.595.86--
VinVL+L64.8582.5994.04.5949.1984.9196.62VinVL+L: Enriching Visual Representation with Location Context in VQA
KU55.072.0983.475.2939.9284.6696.34--
BottomUp49.7466.6478.715.9834.8384.5796.18Bottom-Up and Top-Down Attention for Image Captioning and Visual Question Answering
graphRepresentation, Single57.8974.5485.455.7343.1984.9996.4--
GRN61.2278.6990.316.7745.8185.4396.36Bilinear Graph Networks for Visual Question Answering-
total1456.9574.6287.715.8141.3684.5795.98--
TESTOVQA00760.1876.9789.655.2945.3684.4796.33--
NSM single (updated)63.1778.9493.253.7149.2584.2896.41--
LW55.6572.8689.189.6940.4685.2796.33--
GM6_9_2_train56.9674.9785.127.1341.0684.8596.38--
0 of 127 row(s) selected.