VQA: Visual Question Answering

ICCV 2015 Aishwarya AgrawalJiasen LuStanislaw AntolMargaret MitchellC. Lawrence ZitnickDhruv BatraDevi Parikh

We propose the task of free-form and open-ended Visual Question Answering (VQA). Given an image and a natural language question about the image, the task is to provide an accurate natural language answer... (read more)

PDF Abstract
TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK RESULT BENCHMARK
Visual Question Answering COCO Visual Question Answering (VQA) abstract 1.0 multiple choice LSTM + global features Percentage correct 69.21 # 3
Visual Question Answering COCO Visual Question Answering (VQA) abstract 1.0 multiple choice LSTM blind Percentage correct 61.41 # 4
Visual Question Answering COCO Visual Question Answering (VQA) abstract 1.0 multiple choice Dualnet ensemble Percentage correct 71.18 # 2
Visual Question Answering COCO Visual Question Answering (VQA) abstract images 1.0 open ended Dualnet ensemble Percentage correct 69.73 # 2
Visual Question Answering COCO Visual Question Answering (VQA) abstract images 1.0 open ended LSTM + global features Percentage correct 65.02 # 3
Visual Question Answering COCO Visual Question Answering (VQA) abstract images 1.0 open ended LSTM blind Percentage correct 57.19 # 4
Visual Question Answering COCO Visual Question Answering (VQA) real images 1.0 multiple choice LSTM Q+I Percentage correct 63.1 # 8
Visual Question Answering COCO Visual Question Answering (VQA) real images 1.0 open ended LSTM Q+I Percentage correct 58.2 # 11
Visual Question Answering COCO Visual Question Answering (VQA) real images 2.0 open ended HDU-USYD-UNCC Percentage correct 68.16 # 1
Visual Question Answering COCO Visual Question Answering (VQA) real images 2.0 open ended DLAIT Percentage correct 68.07 # 2

Methods used in the Paper


METHOD TYPE
🤖 No Methods Found Help the community by adding them if they're not listed; e.g. Deep Residual Learning for Image Recognition uses ResNet