We were brainstorming ideas as the hackathon began and one of our team members suggested that we create a program that could identify a face for someone's contact information. We hoped that it would be a useful tool for the disabled, especially in the blind community. If program could be opened on the person's phone, they would be able to discreetly tell who was approaching them without having to ask. This would be especially helpful for people who were not blind at birth as it would help them adjust to their new way of life.
What it does
Iris is a simple facial recognition program that was created to aid the blind in identifying faces. If a face is recognized, the program will display the name of individual on the screen. This program could revolutionize accessibility for the blind community by allowing them to understand who they're speaking to with a simple click of a button.Iris is a web-based site which can be accessed on the phone. Using a web short-cut on their phones, the blind can access the website easily. The homescreen of the website contains a camera setup that allows a photo to be taken. Then with their own personal data base, the application runs the photo through the database, looking for similarites from previously submitted entities. If the photo was not entered into the database or is not recognized, a person may enter the photo into the database, adding a name with or without notes.
How we built it
First we modified a face detection API to be able to match faces to names in an existing database of contacts. The python program then displays the name of the person that matches the existing photo in the database. For the website itself, we created a download link connected to the logo on the home page. It links directly to the python program for face recognition. The other pages include a mockup of a contact form and information about us and the program.
Challenges we ran into
For the website itself, we ran into challenges using PHP and MySQL for storing user-inputted data from the contact form. We were also hoping to create a way for the user to upload their own photos to compare faces to but we were unable to make the database work. We also had trouble linking the python and HTML files together using programs such as Flask, CGI, or Brython (Browser Python) which is why we had to use a download link instead.
Accomplishments that we're proud of
We are proud of having a facial recognition software that works somewhat. Additionally, many of us have very little experience with programming, but by working with mentors and workshops, we were able to overcome this obstacle. We were lso able to come up with our own logo.
What we learned
We learned a lot about working as a team and communicating effectively. We had to work together to try to connect the Python and HTML code. Although we didn't succeed in the end, we learned a lot about how to manage time and find resources online to help us learn how to code.
What's next for Iris
We hope finish the Iris project and help the blind community. In the future, we may expand the interface to apps, both iOS and Android.