This is a Web Application which matches your Query image to your Reference Database and then retrieving the image from the database with the highest similarity.
You can find the project Live here.
-
Upload two images (target) with which you will compare the query (from webcam).
-
Once uploaded, use face-api library to crop the uploaded images.
-
Prompt a message to re-upload if more than one face is in the image. Imagine a police database where officers upload a single front-facing image.
-
Ask User to add a query image.
-
Use the face-api to crop the face from the query image and display them.
-
If query image has multiple faces, let user choose one of those.
-
Now use SkylarkLabs Face Recognition https endpoint(given below) to display the face with most similarity. [https://skylark-test-site.scrollhelp.site/FMAD/Face-Recognition-API.81821697.html]
-
The API matches two cropped faces and returns the similarity.
NOTE : Reference Image / Target Image is the suspect whom you want to find out in Query Image.
-
Open terminal / command prompt and Clone the project using
$ git clone https://github.com/mnamegaurav/facematch.git
-
Install docker and run this command in your system:
$ sudo docker build -t facematch:latest .
-
Run this project:
$ sudo docker run --name facematch -d -p 3000:3000 facematch:latest
-
Now you are good to go, Open browser and go to
http://localhost:3000
.
-
Open terminal / command prompt and Clone the project using
$ git clone https://github.com/mnamegaurav/facematch.git
-
Install
npm
dependecies.$ npm install
-
Signup for the API key of SkyLark in and paste it in the
.env.example
and rename this file to.env
. -
Start the development server:
$ npm start
-
Now you are good to go, Open the browser and go to
http://localhost:3000
.