Nueral network model deployed using Heroku and Flask to classify Christmas Images
Model is trained on dataset with 8 different classes of christmans images namely:
- christmas cookies
- christmas presents
- christmas_tree
- fireworks
- penguin
- reindeer
- santa
- snowman
Download test folder from here. It consists of test.py python file and sample images which can be used to check model's prediction.
Add image to this folder which needed to be classified in mentioned classes. Image should be suitable to 8 different classes.
Minor change in test.py file:
import requests
resp = requests.post("https://christmas-model-b04.herokuapp.com/predict", files={'file':open('santa.png', 'rb')})
print(resp.text)
Change santa.png with the name of your image as image_name.extension
Note: This model will only allow .jpg, .jpeg and .png formats of image. Please consider it before.
Move to test folder by typing cd test in terminal
Run test.py file using following command in terminal python test.py
This classification challenge was organized as a part of Deep Learning subject at Universität Siegen by Computer Vision Group via Kaggle platform.
One can easily find mentioned challenge, training dataset and its leaderboard for year 2022/23 HERE