Fashion-MNIST Image Classification
This repository is a collection of TensorFlow machine learning models for Fashion-MNIST image classification benchmarking.
Available models
Convolutional Neural Networks
Recurrent Neural Networks
How to run
python fashion-mnist.py -m [model] -s [model_name] -p [pretrained_model]
Example to run and save a model:
python fashion-mnist.py -m LeNet -s pretrained_LeNet
Example to load and run a pre-trained model:
python fashion-mnist.py -m LeNet -p pretrained_LeNet
Run without arguments to get a list of available models.
Results
Based on training with 60,000 images and testing with 10,000 images from the Fashion-MNIST dataset (sorted by accuracy).
| Model | Type | Accuracy | Precision | Recall |
|---|---|---|---|---|
| ResNet | CNN | 0.9111 | 0.9133 | 0.9096 |
| AlexNet | CNN | 0.9060 | 0.9104 | 0.9010 |
| LSTM | RNN | 0.8959 | 0.9047 | 0.8884 |
| LeNet | CNN | 0.8882 | 0.8960 | 0.8839 |