Visual Question Generation on VQA 1.0
65.1BLEU-1MDN-Joint
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| MDN-JointIntegration module=Joint2018.08 | 65.1 | 22.7 | 52 | 53.1 | |
| MDN-HadamardIntegration module=Hadamard2018.08 | 61.7 | 16.7 | 50.1 | 29.3 | |
| MDN-AdditionIntegration module=Addition2018.08 | 61.7 | 18.3 | 50.4 | 42.6 | |
| MDN-AttentionIntegration module=Attention2018.08 | 60.7 | 16.7 | 49.8 | 33.6 | |
| MaxAggregation strategy=Max2018.08 | 59.4 | 17.8 | 49.3 | 33.1 | |
| Caption OnlyInput=Caption Only2018.08 | 57.1 | 15.5 | 36.6 | 30.5 | |
| Image OnlyInput=Image Only2018.08 | 56.6 | 15.1 | 40 | 31 | |
| SampleAggregation strategy=Sample2018.08 | 38.8 | 12.7 | 34.2 | 13.3 |