The research paper can be accessed from here https://arxiv.org/pdf/1512.03385v1.pdf
The dataset used here is CIFAR10
The accuracy of the model are as folows:
MODEL | ACCURACY |
---|---|
ResNet18 | 91.89 |
ResNet34 | 92.77 |
ResNet50 | 86.96 |
RedNet101 | 85.56 |
ResNet152 | 84.65 |
PyTorch 1.0+ and Python 3.6+
Implement the same on other datasets like ImageNet, MNIST, COCO. And try to improve the accuracy of the models generated