Authors:

Major: Computer Engineering

Date Created:

Not specified

Course Title:

Professor:

Not specified

About Paper:

Multi-Agent Classification using Implicit Models Classifying agents using sparse, noisy sensor networks becomes significantly more challenging when you have multiple agents moving through an environment simultaneously. As sensor observations from different agents become interweaved, a data association problem of determining which data belongs to which agent becomes an challenge that must be solved alongside classification. When false positive detections, missed observations, and measurement noise are accounted for, additional uncertainty is added to the problem. This thesis addresses this problem of multi-agent classification by integrating implicit agent models with Markov Chain Monte Carlo (MCMC) sampling techniques to jointly associate observations with individual agents and classify their behaviors in accordance to their movement behavior. These implicit models define agents through simulation rather than explicit probability distributions, allowing for a flexible representation of complex movement patterns. The approach described in this thesis, uses edit distance calculations to compare observed sensor sequences against simulated trajectories generated from the implicit models, while MCMC sampling explores the space of possible observation-to-agent assignments that are then used to fully classify our agents. The results demonstrate the feasibility of extending implicit model classification frameworks to multi-agent scenarios and have potential applications in surveillance systems, wildlife monitoring, and other fields where multiple entities must be tracked and classified using a distributed sensor network. In doing so, this thesis presents a promising approach for multi-agent tracking in environments where sensor data is sparse and imperfect. Poster Session 5 3:30 PM-4:30 PM CT Room C Poster #10: Lillian Mitchell Major: Computer Engineering

Source:

Texas A&M University / 2026

Topics:

No topics listed

Co-authors:

Major: Computer Engineering