umbertogriffo / fashion-mnist-cnn-keras

Zalando's article images Recognition using Convolutional Neural Networks in Python with Keras.

Zalando's article images Recognition using Convolutional Neural Networks in Python with Keras

• Author: Umberto Griffo
• Twitter: @UmbertoGriffo

Software requirements

* Python 3.6, TensorFlow 1.11.0, Keras 2.2.4, numpy, matplotlib, scikit-learn, h5py

Training

execute fashion_mnist_cnn.py

Preprocessing

Normalization

Cross Validation

5-fold cross-validation

CNN configuration

The network topology can be summarized as follows:

- Convolutional layer with 32 feature maps of size 5×5.
- Pooling layer taking the max over 2*2 patches.
- Convolutional layer with 64 feature maps of size 5×5.
- Pooling layer taking the max over 2*2 patches.
- Convolutional layer with 128 feature maps of size 1×1.
- Pooling layer taking the max over 2*2 patches.
- Flatten layer.
- Fully connected layer with 1024 neurons and rectifier activation.
- Dropout layer with a probability of 50%.
- Fully connected layer with 510 neurons and rectifier activation.
- Dropout layer with a probability of 50%.
- Output layer.

Results

I evaluated the model using the 5-fold cross-validation on 60,000 examples divided into train and test.

Accuracy scores: [0.92433, 0.92133, 0.923581, 0.92391, 0.92466]

Mean Accuracy: 0.923567

Stdev Accuracy: 0.001175

I ran a new learning from scratch on 60,000 examples and then I evaluated test accuracy on the test set of 10,000 examples.

Final Accuracy: 92.56%

The following picture shows the trend of the Accuracy of the final learning:

References

• [1] Fashion-MNIST https://github.com/zalandoresearch/fashion-mnist