Here we aim to reproduce the MNIST results from this paper via an implementation in Keras:

Y. Wen, K. Zhang, Z. Li, and Y. Qiao, ‘A Discriminative Feature Learning Approach for Deep Face Recognition’, in Computer Vision – ECCV 2016, 2016, pp. 499–515.

Content:

• Results

• Links

Results

First we train a classifier using cross-entropy loss alone (basic_MNIST.py):

We then add the 'center loss term' (centerloss_MNIST.py):

Links

The original authors implement in Caffe:

https://github.com/ydwen/caffe-face

Some other repos which are trying to do similar:

https://github.com/shamangary/Keras-MNIST-center-loss-with-visualization (Keras)
Warning: Not sure if this is the same method as the paper presents

https://github.com/EncodeTS/TensorFlow_Center_Loss (TensorFlow)
Warning: Readme visualizations are on training set

https://github.com/jxgu1016/MNIST_center_loss_pytorch (PyTorch)
Warning: Readme visualizations are on training set