EMPath at Affectiva: A case study
How we created a customized AI program that went global in less than 3 years
About Affectiva
Affectiva is an MIT Media Lab spin-off that created the Emotion AI category and is leading the space. Built on deep learning, computer vision, speech science, and real-world data, Affectiva’s technology can detect nuanced human emotions, complex cognitive states, activities, interactions, and objects people use. Affectiva’s technology is used by 25% of the Fortune Global 500 companies to test consumer engagement with ads, videos, and TV. Affectiva's AI is also used by leading car manufacturers to understand what is happening with people in a vehicle, enhancing road safety and unlocking more personalized and comfortable consumer experiences.
EMPath 2020 by the numbers
52 Trainees
50% Female
4 Different countries: United States, Canada, Egypt, India
20% with less than 1 year of coding experience
A pathway to learning about AI
In 2017, Affectiva embarked on a small AI training program called EMPath, developed and led by Taniya Mishra, who at the time held the roles of Lead Speech Scientist and Director of AI Research at Affectiva. Beginning with just one student in 2017, EMPath eventually grew into a 52-student program in 2020, run completely virtually with participants from four different countries.
Students in EMPath could choose from seven different learning curricula that develop the technical skills needed for building AI products to improve human-machine interactions. Every student also learned about Affective Computing, a branch of AI that pertains to understanding, simulating, and reacting to human emotions — Affectiva’s core business focus. The program customization allowed Affectiva to support its own R&D goals while training and mentoring the next generation of AI minds.
“I've learned so much — not limited to machine learning and coding, but also about how to market myself in the tech world, the inclusivity of tech workplaces, and how to lead a happy, successful life in general.”
Ashna, EMPath 2020 student
"I loved this experience, and I love how it was designed to empower students. As a mentor, I feel like I learned so much as well. I loved being part of this energetic learning community!"
Christopher, EMPath 2020 mentor
Led by interest, guided by mentors
EMPath 2020’s 5-week program began with students engaging in self-directed learning, along with one-on-one and group mentoring from over 40 different mentors. Students also attended 16 online expert-led talks on professional development, entrepreneurship, marketing, and more.
Turning skills into actions
At the halfway point in the program, students were organized into teams and tasked with applying their new learnings in a Makeathon. The challenge was to build a working AI prototype that applied affective computing to improve human-machine interactions in four application areas: tele-health, automotive, IoT and video conferencing. On the last day of the program, teams had to pitch their final projects to a judging panel of industry leaders as well as a large virtual audience. The results were astounding!
See the winning projects from EMPath 2020, a fully virtual AI training program, developed and led by Dr. Taniya Mishra and hosted by Affectiva.
The winning projects
1st place
A smart mirror that understands emotions in real time and converses with you.
2nd place
A Zoom integration for video-conferencing that showed live updates of audience emotions and engagement to the presenter.
3rd place
An emotion-enabled therapeutic platform to help children with learning difficulties receive tele-therapy.
Building mindful leaders
What started with one mentor, Taniya, in 2017, grew to over 40 mentors in 2020, including Affectiva staff and graduate students from top universities. This was not incidental. EMPath wasn’t just about training for technical AI skills; it was also about building the next generation of leaders within the company and cultivating a pipeline of potential future hires.