Face-recognition-tfjs

The repository of the Face Recognition TFjS article on deploying models.

Downloading Requirements

To start, you will require

• Chrome Web Server
• Brackets (or any other editor) if you want to play with code (optional)

How to download Chrome Web Server (200 OK)

Please download the repository before doing this

1. Go to Web Server for Chrome
2. Install the web server, and open it.
3. Press the CHOOSE FOLDER button and select the folder in which you have the files of the repository.
4. Press the link in the Web Server application (for example 127.0.0.1:8887)
5. Select the facerecgn.html file to see the demo

How to download Brackets (optional)

1. Go to the brackets website
2. Select platform and download
3. Configure and use.

Using the demo

As soon as you complete downloading the repo files and Chrome Web Server, you will have the facerecgn.html file open

1. After you permit Webcam access, please enter names in the three input fields and submit them.
2. With that, you can start adding images to the dataset by first ensuring that (only that) person is seen in the webcam, and then press on his/her name's button.
3. Repeat till you have about 50 pictures of all three people.
4. Now, open the console by going to the three dots in the top right of your Chrome window, hover on More tools and then press Developer tools.
5. Now you can press the Train Network button. You should see the losses being printed in the console after some time.
6. After Training over. You are good to go is printed in the console, you can press Start Predicting, and see the predictions.