mxnet-center-loss
所属分类:数值算法/人工智能
开发工具:Python
文件大小:1515KB
下载次数:0
上传日期:2017-11-08 05:07:26
上 传 者:
sh-1993
说明: mxnet中心丢失,mxnet Gluon中心丢失的实现:一种用于深度人脸识别的判别特征学习方法
(mxnet-center-loss,MxNet Gluon Implementation of Center Loss: A Discriminative Feature Learning Approach for Deep Face Recognition)
文件列表:
center_loss.py (887, 2017-11-08)
main.py (5790, 2017-11-08)
models (0, 2017-11-08)
models\LeNet.py (1968, 2017-11-08)
models\__init__.py (36, 2017-11-08)
output (0, 2017-11-08)
output\center-test.gif (170047, 2017-11-08)
output\center-test.png (217400, 2017-11-08)
output\center-train.gif (203732, 2017-11-08)
output\center-train.png (223385, 2017-11-08)
output\curves.png (28652, 2017-11-08)
output\softmax-test.gif (156816, 2017-11-08)
output\softmax-test.png (240548, 2017-11-08)
output\softmax-train.gif (145503, 2017-11-08)
output\softmax-train.png (203066, 2017-11-08)
requirements.txt (35, 2017-11-08)
utils.py (2250, 2017-11-08)
# MxNet implementation of the paper: A Discriminative Feature Learning Approach for Deep Face Recognition
## Requirements
```
pip install -r requirements.txt
```
## Training
1. Train with original softmax
```
$ python main.py --train --prefix=softmax
```
2. Train with softmax + center loss
```
$ python main.py --train --center_loss --prefix=center-loss
```
## Test
1. Test with original softmax
```
$ python main.py --test --prefix=softmax
```
2. Test with softmax + center loss
```
$ python main.py --test --prefix=center-loss
```
## Image
Comparison Accuracy curve:
### Softmax
Training:
Testing:
### Softmax + Center Loss
Training:
Testing:
近期下载者:
相关文件:
收藏者: