ResNet 18 Caffemodel On ImageNet Save
ResNet-18 Caffemodel @ilsvrc12 shrt 256 with Top-1 69% Top-5 89%
Project README
ResNet-18-Caffemodel-on-ImageNet
Accuracy
We reported the test accuracy on ImageNet (ILSVRC2012 Validation Set).
| DataSet | Top-1 | Top-5 | Loss |
|---|---|---|---|
| Both256 | 67.574% | 88.1001% | 1.33896 |
| Shrt256 | 69.0801% | 89.0321% | 1.2711 |
About shrt 256
Augmented training and test samples:
This improvement was first described by Andrew Howard [Andrew 2014]. Instead of resizing and cropping the image to 256x256, the image is proportionally resized to 256xN(Nx256) with the short edge to 256. Subcrops of 224x224 are then randomly extracted for training.
Model Link
Open Source Agenda is not affiliated with "ResNet 18 Caffemodel On ImageNet" Project. README Source: HolmesShuan/ResNet-18-Caffemodel-on-ImageNet
Stars
190
Open Issues
2
Last Commit
4 years ago
Repository
Open Source Agenda Badge
