FIT’s Siddhartha Bhattacharyya is working to improve computer vision-based autonomous systems, such as his lab’s Advanced Line Identification and Notation Algorithm. His team will use data they collect from airports and train their models to attempt to drive an aircraft autonomously. They are capturing images of runways from different angles and weather and lighting conditions so the ALINA model learns to identify the most accurate course regardless of conditions.

The team will explore both transfer learning-based approaches and few-shot learning approaches.  FSL is a technique that teaches a model to generalize information with just a few data samples instead of the massive datasets used in transfer learning. “That would help us reduce the time and cost of data collection as well as time spent labeling the data.

• Siddhartha Bhattacharyya