https://github.com/shunk031/chainer-center-loss

Implementation of Center Loss in Chainer
https://github.com/shunk031/chainer-center-loss

center-loss chainer chainerv2

Last synced: 3 months ago
JSON representation

Implementation of Center Loss in Chainer

• Host: GitHub
• URL: https://github.com/shunk031/chainer-center-loss
• Owner: shunk031
• Created: 2017-07-19T15:33:36.000Z (over 8 years ago)
• Default Branch: master
• Last Pushed: 2018-03-02T05:28:04.000Z (over 7 years ago)
• Last Synced: 2025-04-30T02:38:22.837Z (6 months ago)
• Topics: center-loss, chainer, chainerv2
• Language: Python
• Homepage:
• Size: 222 KB
• Stars: 4
• Watchers: 2
• Forks: 1
• Open Issues: 1
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# Chainer Center Loss

Implementation of [Center Loss](https://link.springer.com/chapter/10.1007/978-3-319-46478-7_31) in Chainer.

## Requirements

- Python 3.5.1

- Chainer 2.0.0

- CuPy 1.0.0 (if use GPU)

- Matplotlib

## How to train

* with CPU and use center loss

```shell

python train_mnist.py --batchsize 32 --epoch 20 --centerloss

```

* with GPU and use center loss

```shell

python train_mnist.py --batchsize 32 --epoch 20 --gpu 0 --centerloss

```

## Visualize the result

* MNIST, Softmax Loss + Center Loss  

  The white dots denote 10 class centers of deep features.

![Figure 1](https://raw.githubusercontent.com/shunk031/chainer-center-loss/master/img/visualize_deep_feature_with_center_loss.png)

* MNIST, only Softmax Loss

![Figure 2](https://raw.githubusercontent.com/shunk031/chainer-center-loss/master/img/visualize_deep_feature_without_center_loss.png)

## Reference

- [Wen, Yandong, et al. "A discriminative feature learning approach for deep face recognition." European Conference on Computer Vision. Springer International Publishing, 2016.](https://link.springer.com/chapter/10.1007/978-3-319-46478-7_31)