IRIS.AI

IRIS.AI Software Logo

Inspiration

Our two main focuses have been:

Helping neurodivergent people: People with neurodivergences, like autism, that struggle with identifying other people's emotions and language expressions like irony.
Assisting professionals: Psychologists and/or psychiatrists that need to identify a patient's emotions in out-of-normal conditions like children or criminology, detectives during interrogation, company recruiters, etc.

What it does

IRIS.AI is an emotion and behavioral pattern recognition software, implementing the use of AI face and emotion recognition through camera or recordings in different types of hardware (glasses, Meet live sessions, etc).

How we built it

In the beggining we made 2 different teams to split up the workload evenly. One team was in charge of the directing and editing of the video presentation and the other was in charge of creating a working protoype. To create the prototype we started by researching for APIs and adapting the code for our needs. This brought different problems as we were working with limited hardware, but in the end we managed to fixed it after hours of research. At the end of the day, we were able to use our working prototype in our own video, showing how much we improved the code, even solving the initial crash problem that had us stuck for several hours.

Challenges we ran into

The main challenges we encounter was the implementation of an API, for it to be useful with AR Glasses and helpful to neurodivergent people. Another challenge we had to face was video editing in a professional way and merging it with the working prototype.

Accomplishments that we're proud of

Making the Hugging Face's API work and using it in the editing of the video presentation made our idea come to life, which is what we are most proud of.

What we learned

Through this short-length process we've learned a lot, about programming, team work, organization skills, etc. But what we've learned most about is:

1. Video Editing and Direction:

After watching references from the tech world, we were inspired to create a video presentation resembling those seen in tech company’s hardware presentations. Looking for a free-to-use platform that could handle the project’s demands was time-consuming, but in the end, we were able to find the correct service provider to begin the editing process. With the inspirations we had we were able to direct and edited it to accomplish a professional looking video presentation.

2. Implementation of the API:

We knew the implementation of an AI API was necessary to create a working prototype. After going through several APIs (Google, Microsoft, etc.), we were convinced by Hugging Face’s API as its instructions were the most adaptable to our needs and vision. To put it to use we had to learn how to use python, library installation, environment creation and storage optimization.

What's next for IRIS.AI

The future of IRIS.AI is to implement a series of subscription plans to make a self-sustaining development out of this project, to let us keep improving on it further along the way:

Free Plan: The basic IRIS.AI experience will help people recognize human emotions, facial expressions, and behavioral patterns, without storing data from encounters. Free of charge.
Premium Plan: For a better IRIS.AI experience, a personal database associated with the client’s phonebook allows IRIS.AI to adapt its predictions based on their acquaintances' emotions and expressions. Monthly subscription: $19.99. Yearly subscription: $199.99.
Business Plan: For businesses that want their employees to have access to the IRIS.AI software, allowing for multiple users in one subscription. This plan also offers incremented database storage, personalized advanced analysis, and recommendations for efficient hardware usage. It also includes direct access to our technical support. Monthly subscription: $599.99. Yearly subscription: $5999.99.