Reproduction of the architectures and results of some of the most significant papers that came out of the ImageNet challenge.
This project uses Python 3.7.5. Setup your virtual environmnet of choice then install with the below
pip install -e .| Paper | Year | Val Top-5 % | Reproduction |
|---|---|---|---|
| AlexNet | 2012 | 83.00 | |
| ZFNet | 2013 | 84.00 | |
| GoogLeNet | 2014 | 89.93 | |
| VGGNet | 2014 | 92.00 | |
| ResNet | 2015 | 95.51 | |
| Xception | 2016 | 94.50 | |
| SENet | 2017 | 97.75 |