object_classification_with_MNIST-fashion

Item Classification with MNIST fashion dataset. Images are of size 28 * 28 size.
Download dataset from here : https://www.kaggle.com/zalando-research/fashionmnist
Items are namely :

ID	Item Name
0	T-shirt/top
1	Trouser
2	Pullover
3	Dress
4	Coat
5	Sandal
6	Shirt
7	Sneaker
8	Bag
9	Ankle boot

Here is image of showing full dataset in a glance.

Requirements

1. Tensorflow-gpu v1.3.0
2. Keras v2.1.3
3. Pandas
4. Numpy v1.14.0
5. sklearn v0.19.1
6. matplotlib
7. PIL (python image library) v4.2.1
8. OpenCV (cv2) v3.3.0

Before usage

Neural network only give best results when we give similar images on which it is trained in testing phase.
So, I am showing some images on which it is trained.
You should use this model with "similar" images. Not on completly different images.

1. T-shirt/top :

2. Trouser :

3. Pullover :

4. Dress :

5. Coat :

6. Sandal :

7. Shirt :

8. Bag :

9. Ankle Boot :

Usage

Run command

python test.py ImagePath

In short give path of test image as command line argument.