Scouter | Devpost

Start menu of the app
Scanning a recognized face
Data displayed in real time on the screen

Inspiration

Black Mirror: https://s30.postimg.org/vq7gihvu9/nodedive_Screen.jpg Silicon Valley: Scanning people with an app to see their credentials. How to differentiate “important” people from everyone else from an investment standpoint.

What it does

Our Facial Recognition iOS application has 2 functions: -It allows people to “make an account” by scanning their own face and inputting their credentials and basic information -It allows people to scan others from across the room and seek out their information at a press of a button.

How we built it

We utilized a custom camera built into an IOS app. Through apples built in facial recognition class we analyzed faces and called Microsoft's cognitive API to create a person profile for that individual that would store the facial features. Then through another camera, we scan the room and find the faces. These faces are then cropped and sent to the server and matched with the most similar face features we have stored. That person's data is then displayed live on screen on top of the scanned individual.

Challenges we ran into

API request limits to Microsoft cognitive servers on a free plan. Combining live video feeds from IOS camera and cropping out face recognition into JPG files that are then sent to Microsoft cognitive API seamlessly in the background to then retrieve data.

Accomplishments that we're proud of

Being able to manipulate Microsofts cognitive API that is usually only seen on desktops into a mobile application that is lightweight and usable on the go.

What we learned

How to manipulate Microsoft's cognitive API. How to train the neural net to identify faces uploaded from a mobile application to identify people. How to combine a live video and cropped pictures and send them to a server in real time without interrupting the video feed.

What's next for Scouter

To allow users to sign up using Facebook or Linkedin so their information can be scraped and ready to go immediately when they sign up. And allowing to scan multiple faces at once and display that data on screen at the same time.

Built With

Submitted to

Hack @ UCI

Created by

I worked on a web crawler using jquery to scrape websites looking for person's information.
We used this information to add it to the scanned face profile.
I also helped in implementing Microsoft's API for facial recognition part of the app.

Sasha Mirabi
I worked on researching the technology to use for facial recognition, and I learned the API to support the programmer of the app.

Also I helped flesh out the idea.

Seung-Li Maeda
I worked on implementing a Facebook api to scrape people's profiles to display people's information when scanned. I also worked on implementing the Microsoft API to scan people's faces and store them in a database. I then used built in Microsoft functions to compare scanned photos to photos in the database.

Ryan Ghavimi
Mark Fernandez
Mauro Serrano