Mechanical Engineering sophomore Takara Truong took an impressive second place in the October 27, second annual, 24-hour HackTheU competition, held at the Tower at the University of Utah’s Rice-Eccles Stadium. A “hacker” and “disruptor” can be positive descriptors for tech savvy creatives who come together for hackathons, competitive events that combine communal brainstorming and creative collaboration.
Utah’s largest hackathon, HackTheU, attracted more than 270 college students across several states, who participated in the coding marathon. The participants competed for cash and prizes in several tracks: machine learning, fintech (FinX), games for health, and an open option. Additionally, HackTheU included keynote speakers and dozens of mentors from local and national companies.
From Salt Lake City, Truong is interested in being a robotics engineer and is currently a member of the Advanced Energy Innovations Lab directed by mechanical engineering assistant professor Roseanne Warren. His robotics curiosity naturally lead him to the HackTheU competition. Choosing to compete in the category of machine learning, Truong said, “I successfully used transfer learning on a convolutional neural network (CNN) to understand my preferences in female physical traits.” CNNs are a category of neural networks that have proven very effective in identifying faces, objects and traffic signs apart from powering vision in robots and self-driving cars.
Truong built a rig equipped with an Arduino, webcam, servo motor (a rotary actuator that allows for precise control of angular position), that was used to communicate with the neural network. Lastly, he programmed his project to autonomously swipe right or left based on Tinder profile images (taken from his phone by the webcam) according to the AI’s standard of beauty that he had taught it.
“During the competition, the project performed over 4000 swipes in about four and a half hours, of which 60 percent were considered “swipe left” and the rest “swipe right” with about 85 percent accuracy taken from mini batches of 20 swipes,” explained Truong.
“Participating in the competition was a great opportunity to learn how I could utilize machine learning in my own coding. The workshops taught me how I could make the most out of my CNN which overall helped with the accuracy of my project. I was pleasantly surprised when judges and other competitors came to my work area to video my project. One judge commented that my project was another small step for mankind. I greatly enjoyed the experience and will participate next year with a goal of winning first place.”
About HackTheU:
Founded in 2016 by University of Utah student Johnny Le, HackTheU is Utah’s largest hackathon.
“I created HackTheU to bring software innovation and opportunity to students in the region,” said Le. “Creating a tangible solution from your own idea over the course of a weekend is something truly unique that I hope will soon help hundreds and thousands of students to grow.”
Over the course of 24 hours, teams of five students from across the state and nation develop solutions to various prompts through augmented reality, virtual reality, IoT, and many other kinds of applications. The competition, also includes workshops for students on various tech topics presented by local industry leaders as well as mentoring, networking and prizes. Learn more at hacktheu.com.