Image Classification using Tensorflow.js library, could be tested by this link
Simple project for image classification using MobileNet library with tensorflow.js.
What happens:
...
// load image from your disk
// look to this function for understanding how to do it -> https://github.com/MrCrambo/Image-Classification-TFJS/blob/master/static/js/main.js#L12
...
// after loading image to the html load mobilenet model
// do it only one time for saving user time if he will decide to classify other images
net = await mobilenet.load();
...
// then pass your image to classify function and get 3 most possible results
const result = await net.classify(your image element);
...
// then show results
// that's it!In case you will try to build it on your way, just remove index.html file from sourse and leave it only in templates folder, then follow instructions below:
- Clone the repo
$ git clone https://github.com/MrCrambo/Image-Classification-Flask.git- Install Flask
$ pip install flask- Run with Python
$ python app.py- Test it in localhost:5000
- Use your pre-trained model
- Use pre-trained models like
NASNet, etc. - Change UI/UX design
- Recognition of car brands
- Finding facebook profile by only one photo
- Country or place recognition by photo
