Inspiration

In the harsh world of networking, keeping track of every new face is daunting. Do you remember the last time you were introduced to someone new, perhaps a recruiter, and immediately both initiated an engaging conversation? How optimistic did you feel leaving that conversation? Furthermore, how disappointed did you feel when you realized you could not remember their name? That’s where Your Name Is comes into play 一 this is a solution to the endless embarrassment caused by a short memory. This real-time identification system will provide all the necessary information to maintain new acquaintances and remove this setback from the

What it does

This web app is a mobile device-compatible application that identifies new acquaintances using artificial intelligence with a significant focus on facial classification. The user simply opens the application on their mobile device to identify a new person and points the camera-like interface toward them. With the help of facial recognition, Google Cloud and CckroachDB, the new person will be identified in relation to their Linkedin profile. The UI will then present a pop-up containing the new person’s name, interests, position, and profile image.

How we built it

Our backend links to an open-source A.I. facial recognition API. The frontend interface featuring a camera was created using react. These images are pulled from the frontend are analyzed using the facial recognition API. The information which associated to the recognized figure, is stored in a CockroachDB cluster which is pulled from the database in google cloud. We principally used Javascript and node.js to connect our front and backend.

Challenges we ran into

We faced many challenges in selecting a facial recognition API. Many existing APIs mistake two different people as highly similar, and implementation bugs were prominent because of poor API documentation. We equally faced a challenge in uploading image files in a way that can be queried, but this was resolved by uploading images to the cloud using links and reserving the string variable type as the column variable.

Cockroach presented some difficulties, particularly concerning our development timeframe. The learning curve was prominent and establishing a cluster connection to be queried proved difficult with limited mentorship. Additionally, we saw a timely setback in the camera implementation on the front end. We began our development using react native and ultimately decided that react implementation using an enhanced web application would be more efficient.

Accomplishments that we're proud of

As a team, one of our greatest feats was coming to a working understanding of Cockroach using the resources at hand. We also are very enthusiastic about our many educational attempts and successes using facial recognition APIs. Create an enhanced mobile application using mostly technologies unfamiliar to us as a group is our greatest accomplishment this weekend and we are very excited to have successfully implemented our starting idea.

What we learned

As a team, we progressed significantly in our understanding of the applications and how to use CockroachDB efficiently. We also had the opportunity to explore the implementation of a camera in frontend development, the utilization of different facial recognition APIs, and the difficulties that surround augmented reality. In addition, many members of our team were newly exposed to react.js and to the exploration of the creation of a connection between the front and backend.

What's next for Your Name Is

The next step for Your Name Is is, foremost, obtaining permission to access Linkedin’s API in an entrepreneurial capacity in order to expand our reach, affecting a larger circle of people and networks. This software also has unique conceptual applications that would interact well in the context of an AR headset such as Hololense or Google Glass. Finally, an ambitious goal for Your Name Is is the completion of AI software for facial recognition from scratch to fit our program specifications better and possibly new features to be developed.

Built With

Share this project:

Updates