jinguang-dong/cat_or_dog-kaggle-vgg19-tensorflow

A simple Tensorflow code for fine-tuning VGG-19 to solve 'cat or dog' task in kaggle.

Cat_or_dog-kaggle-vgg19-tensorflow

A simple Tensorflow code for fine-tuning VGG-16 to solve 'cat or dog' task in kaggle.

Requirements

Tensorflow > 1.3

Python > 2.7

vgg_19.ckpt from Tensorflow slim pre-trained model https://github.com/tensorflow/models/tree/master/research/slim

Data sets from kaggle https://www.kaggle.com/c/dogs-vs-cats-redux-kernels-edition/data

You should place vgg_19.ckpt in 'check_point' file folder and data in the 'data' file folder

Details

Fine-tune the last layer of VGG16 to tell cat or dog images. I extract 12500(30% of train set) validation samples from train set.

Batch size: 32

learning rate: 1e-4

dropout: 0.5

No other settings in order to make a simple initialization.

Run

Train: python train.py

Test and obtain result: python test.py

output

97% accuracy on validation set.