Vision GNN: An Image is Worth Graph of Nodes

Open paper
Benchmark Model Rank Results
image-classification-on-imagenetPyramid ViG-B#385Top 1 Accuracy: 83.7%Number of params: 92.6MGFLOPs: 16.8
image-classification-on-imagenetPyramid ViG-M#454Top 1 Accuracy: 83.1%Number of params: 51.7MGFLOPs: 8.9
image-classification-on-imagenetPyramid ViG-S#558Top 1 Accuracy: 82.1%Number of params: 27.3MGFLOPs: 4.6
image-classification-on-imagenetPyramid ViG-Ti#798Top 1 Accuracy: 78.2%Number of params: 10.7MGFLOPs: 1.7

Rank counts only results with a code link.