https://github.com/shicai/DenseNet-Caffe

DenseNet Caffe Models, converted from https://github.com/liuzhuang13/DenseNet
https://github.com/shicai/DenseNet-Caffe

caffe deep-learning densenet imagenet torch

Last synced: 8 days ago
JSON representation

DenseNet Caffe Models, converted from https://github.com/liuzhuang13/DenseNet

• Host: GitHub
• URL: https://github.com/shicai/DenseNet-Caffe
• Owner: shicai
• Created: 2017-02-14T09:34:01.000Z (over 7 years ago)
• Default Branch: master
• Last Pushed: 2019-08-29T17:45:11.000Z (about 5 years ago)
• Last Synced: 2024-08-01T22:49:53.143Z (3 months ago)
• Topics: caffe, deep-learning, densenet, imagenet, torch
• Size: 12.7 KB
• Stars: 355
• Watchers: 20
• Forks: 274
• Open Issues: 17
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

• awesome-image-classification - unofficial-caffe : https://github.com/shicai/DenseNet-Caffe
• awesome-image-classification - unofficial-caffe : https://github.com/shicai/DenseNet-Caffe

README

        
# DenseNet-Caffe

### Introduction

We manually converted the original torch models into caffe format from https://github.com/liuzhuang13/DenseNet.

For details of these networks, please read the original paper:

- [Densely Connected Convolutional Networks](https://arxiv.org/abs/1608.06993)

### Pretrained DenseNet Models on ImageNet

The top-1/5 accuracy rates by using single center crop (crop size: 224x224, image size: 256xN)

Network|Top-1|Top-5|Download|Architecture

:---:|:---:|:---:|:---:|:---:

DenseNet 121 (k=32)| 74.91| 92.19| [caffemodel (30.8  MB)](https://drive.google.com/open?id=0B7ubpZO7HnlCcHlfNmJkU2VPelE)| [netscope](http://ethereon.github.io/netscope/#/gist/4928834eca7f06261ba0558e0ff63a6a), [netron](http://lutzroeder.github.io/netron?gist=4928834eca7f06261ba0558e0ff63a6a)

DenseNet 169 (k=32)| 76.09| 93.14| [caffemodel (54.6  MB)](https://drive.google.com/open?id=0B7ubpZO7HnlCRWVVdUJjVVAyQXc)| [netscope](http://ethereon.github.io/netscope/#/gist/71335b6e8634327c9b9216619572b3dd), [netron](http://lutzroeder.github.io/netron?gist=71335b6e8634327c9b9216619572b3dd)

DenseNet 201 (k=32)| 77.31| 93.64| [caffemodel (77.3  MB)](https://drive.google.com/open?id=0B7ubpZO7HnlCV3pud2oyR3lNMWs)| [netscope](http://ethereon.github.io/netscope/#/gist/ee808e19615844b8dbc7b13e92abd233), [netron](http://lutzroeder.github.io/netron?gist=ee808e19615844b8dbc7b13e92abd233)

DenseNet 161 (k=48)| 77.64| 93.79| [caffemodel (110 MB)](https://drive.google.com/open?id=0B7ubpZO7HnlCa0phRGJIRERoTXM)| [netscope](http://ethereon.github.io/netscope/#/gist/8fae97d9c66b40b8da443f7f23e9b29b), [netron](http://lutzroeder.github.io/netron?gist=8fae97d9c66b40b8da443f7f23e9b29b)

**Update** (July 27, 2017): for your convenience, we also provide a link to these models on [Baidu Disk](https://pan.baidu.com/s/1gfjD8cF).

### Notes

Due to compatibility reasons, several modifications have been made:

- BGR mean values **[103.94,116.78,123.68]** are subtracted

- **scale: 0.017** is used, instead of the original std values for image preprocessing

- **ceil_mode: false** is used in the first pooling layers ('pool1')