Visual Question Answering On Clevr
Métriques
Accuracy
Résultats
Résultats de performance de divers modèles sur ce benchmark
Tableau comparatif
| Nom du modèle | Accuracy |
|---|---|
| neural-symbolic-vqa-disentangling-reasoning | 99.8 |
| mdetr-modulated-detection-for-end-to-end | 99.7 |
| a-simple-neural-network-module-for-relational | 95.50 |
| transparency-by-design-closing-the-gap | 99.1 |
| interpretable-visual-reasoning-via-induced | 99.4 |
| learning-visual-question-answering-by | 98.8 |
| compositional-attention-networks-for-machine | 98.9 |
| inferring-and-executing-programs-for-visual | 96.9 |
| nesycoco-a-neuro-symbolic-concept-composer | 99.7 |
| language-conditioned-graph-networks-for | 97.9 |
| the-neuro-symbolic-concept-learner-1 | 98.9 |
| question-guided-hybrid-convolution-for-visual | 65.90 |
| explainable-and-explicit-visual-reasoning | 97.7 |
| ddrprog-a-clevr-differentiable-dynamic | 98.3 |
| film-visual-reasoning-with-a-general | 97.7 |