Multiple Choice Visual Question Answering on Visual7W (test)
72.3Accuracy (MC)CTIwBoxes
Evaluation Results
| Method | Links | |
|---|---|---|
| CTIwBoxesImage features=Bottom-up features, Training data=train + validation2019.09 | 72.3 | |
| CFR2021.10 | 71.9 | |
| CTIImage features=Grid features (ResNet-512), Training data=train + validation2019.09 | 69.3 | |
| CTI2021.10 | 69.3 | |
| STLImage features=Grid features (ResNet-512), Training data=train + validation2019.09 | 68.2 | |
| STL2021.10 | 68.2 | |
| BAN2Image features=Grid features (ResNet-512), Training data=train + validation2019.09 | 67.5 | |
| BAN2021.10 | 67.5 | |
| MLPImage features=Grid features (ResNet-512), Training data=train + validation2019.09 | 67.1 | |
| fPMCImage features=Grid features (ResNet-512), Training data=train + validation2019.09 | 66 | |
| fPMC2021.10 | 66 | |
| MCBImage features=Grid features (ResNet-512), Training data=train + validation2019.09 | 62.2 | |
| SANImage features=Grid features (ResNet-512), Training data=train + validation2019.09 | 61.5 |