Parth Ganeriwala is a Ph.D. student in Computer Science at Florida Tech, navigating the intricate world of Transfer Learning and trying to teach machines the subtle art of “been there, done that.” His research dives into how AI models can not only learn new tasks but also generalize that learning—because, apparently, machines have as much trouble with new environments as humans with Wi-Fi passwords. His thesis, “What is Common Knowledge Across Domains?”, is essentially a deep dive into the question: Can we get machines to carry over their skills from one domain to another without acting like it’s their first day all over again?.
In the ASSIST Lab, Parth is working on projects that test the limits of knowledge transfer, applying Transfer Learning to autonomous navigation, IoT security, and cognitive architectures. One of his latest datasets, AssistTaxi, helps autonomous systems identify taxiways accurately, turning a well-worn machine into a cross-environment pro. Parth’s published works cover everything from the system modeling of IoT to lane detection for self-driving cars, each aiming to make AI as adaptable as it is “intelligent.” And when he’s not coaxing AI to recycle its insights, he’s unraveling ancient scripts with few-shot learning, solving sudoku puzzles, or proving his Nintendo Switch is still no match for him.
Transfer Learning, Cognitive Architectures, Computer Vision, Formal Methods, ML, DL, FSL
Published Research Papers:
© Florida Institute of Technology, All Rights Reserved