Lets keep it simple, Using simple architectures to outperform deeper and more complex architectures

Open paper
Benchmark Model Rank Results
image-classification-on-cifar-10SimpleNetv1#129Percentage correct: 95.51
image-classification-on-cifar-100SimpleNetv1#132Percentage correct: 78.37
image-classification-on-imagenetSimpleNetV1-9m-correct-labels#627Top 1 Accuracy: 81.24Number of params: 9.5M
image-classification-on-imagenetSimpleNetV1-5m-correct-labels#731Top 1 Accuracy: 79.12Number of params: 5.7M
image-classification-on-imagenetSimpleNetV1-small-075-correct-labels#880Top 1 Accuracy: 75.66Number of params: 3M
image-classification-on-imagenetSimpleNetV1-9m#916Top 1 Accuracy: 74.17Number of params: 9.5M
image-classification-on-imagenetSimpleNetV1-5m#941Top 1 Accuracy: 71.94Number of params: 5.7M
image-classification-on-imagenetSimpleNetV1-small-05-correct-labels#965Top 1 Accuracy: 69.11Number of params: 1.5M
image-classification-on-imagenetSimpleNetV1-small-075#968Top 1 Accuracy: 68.15Number of params: 3M
image-classification-on-imagenetSimpleNetV1-small-05#981Top 1 Accuracy: 61.52Number of params: 1.5M
image-classification-on-mnistSimpleNetv1#10Percentage error: 0.25

Rank counts only results with a code link.