By Danielle Fretwell
The Francis College of Engineering, Department of Electrical & Computer Engineering, invites you to attend a doctoral defense by Brittany Decker Soboliev on “Adaptive Multi Target Tracking Through Optimal Data Association Technique Selection.”
Candidate Name: Brittany Decker Soboliev
Defense Date: Friday, March 24, 2023
Time: 10 a.m. to noon
Location: This will be a hybrid defense via Zoom and in Ball Hall 302. Those interested in attending should contact the student (Brittany_decker@student.uml.edu) and committee advisor (email@example.com) at least 24 hours prior to the defense to request access to the meeting.
Advisor: Jay Weitzen, Professor, Electrical and Computer Engineering, University of Massachusetts Lowell
- Paul Robinette, Assistant Professor, Electrical and Computer Engineering, University of Massachusetts Lowell
- Orlando Arias, Assistant Professor, Electrical and Computer Engineering, University of Massachusetts Lowell
- James Vogel, Lead Scientist, STR
This dissertation seeks to improve Multi Target Tracking performance in terms of traditional measures of performance without sacrificing latency or computational resource utilization. The motivation of this research is rooted in the increasingly heterogeneous nature of objects and data of interest composing modern tracking scenarios–emphasizing the performance pitfalls of a single, fixed algorithmic approach.
To accomplish this, a classification algorithm is proposed that uses one or more Multi Target Tracking (MTT) measures of performance to select the optimal data association algorithm and corresponding optimal parameterization in approximate real-time to maximize the chosen measures of performance and minimize latency and compute resource utilization. Such an approach to MTT is called the Adaptive Multi Target Tracking Architecture (AMTTA). AMTTA is not intended to be an overly prescriptive approach to tracking, but rather a proposed framework–flexible and moldable when implemented for the intended application and domain. The AMTTA dynamically selects the optimal data association algorithm from an algorithm bank, improving MTT performance for a wider array of tracking scenarios compared to current state of the art.
All interested students and faculty members are invited to attend the online defense via remote access.