Visual Natural Language Inference on V-SNLI (test)
90.4AccuracyMMBT
Evaluation Results
| Method | Links | |
|---|---|---|
| MMBT2019.09 | 90.4 | |
| FiLMBertDescription=FiLM + BERT, Backbone=ResNet-152 features2019.09 | 90.2 | |
| ConcatBertDescription=Concatenation of Bert and Img baselines2019.09 | 90.2 | |
| BertModality=Text-only, Backbone=BERT base-uncased2019.09 | 90.1 | |
| Late FusionDescription=Averages scores of Bert and Img classifiers2019.09 | 90.1 | |
| V-BiMPM2019.09 | 86.99 | |
| ConcatBowDescription=Concatenation of Bow and Img baselines2019.09 | 49.5 | |
| BowModality=Text-only2019.09 | 48.6 | |
| ImgModality=Image-only, Backbone=ResNet-1522019.09 | 33.8 |