We have all tried to find our celebrity lookalikes. We might feel that we look like celebrity X but others might feel we look like celebrity Y. If only there was a way to inject objectivity through a dedicated application.
What it does
The app takes in a public URL to an image of yourself or someone else that you want to try out for. Once you click detect, the image is sent to Clarifai to analyze and return probability scores of celebrities that you look like.
How I built it
The app is a full stack application. It uses React on the frontend for the user interface. It also contains a backend to manage user accounts. The backend was made with Nodejs and Express together with PostgreSQL to store the data. The server and database is hosted on Heroku and the frontend is hosted on GitHub pages. The API validation was also moved to the backend.
Challenges I ran into
This was the first time I built a full stack application. I had no idea how to connect the dots between the frontend, backend and the database. I was constantly running into problems to connect them. Moreover, this was the first time I worked with React, Nodejs, Express & PostgreSQL. So, I spent the first few days to learn these technologies and concepts. Then I built out the full project connecting all the concepts during the hackathon.
Accomplishments that I'm proud of
At the end of this hackathon, I wanted to have a better understanding of full stack development, connecting the frontend to the backend and database. I am happy to say that I have indeed developed a greater understanding and appreciation to these concepts and technologies while also building a real working application!
What I learned
I learned React, NodeJs, Express, PostgreSQL, Full-stack development
What's next for Celebrity Lookalike
Instead of relying on an external machine learning algorithm to do the face detection and processing, I can try to build my own algorithm to do this allowing me to dive into the area of machine learning and computer vision.