AI for Translating American Sign Language (ASL)
The typical translation devices for people who use ASL are gloves that aren't very accurate. The gloves tend to be inaccurate because the translation doesn't only come from peoples' hand gestures, it also comes from their body language, but the gloves don't account for that. However, the even bigger issue with these gloves is that many people can't afford them. Each glove is around $1,500. Tracy and I helped create a translation device for people who use ASL.
Katherine Stevo is a 17-year-old junior at the Buckingham Browne & Nichols School in Cambridge, MA. Her main passion is the intersection of robotics and artificial intelligence. She is a 2018 alumna of AI4ALL, a program that educates young high schoolers about artificial intelligence. At AI4ALL, her team created a translation device utilizing computer vision to assist people who use American Sign Language. She is also a graduate of the Artemis Project and Codebreakers. Her other interest is the ethics of artificial intelligence, specifically mitigating bias, and she spoke about the topic at The Atlantic's Humanity and Tech conference at MIT. In the past couple of months, she won first place at the Code Day Boston Hackathon and also received second place and the people's choice award at MIT's Hack For Inclusion. She is currently working on developing her project from MIT's hackathon, which focuses on assisting visually impaired people to navigate public transportation.