Title: | Enhancing Radar-based Monitoring Systems Through Synthetic Data Generation |
Subject: | Computer science, Embedded systems, Robotics |
Level: | Advanced |
Description: |
This thesis explores the role of synthetic data generation in improving indoor radar-based monitoring systems, particularly for movement analysis and activity recognition. The focus is on using deep learning and generative AI techniques to simulate diverse and realistic movement/activity patterns in an indoor environment for model training, validation, and testing. This approach addresses data scarcity while ensuring privacy and enhancing system performance. By leveraging these advanced methods, the research contributes to developing robust, adaptable, and privacy-preserving monitoring solutions for real-world applications. The work is part of the AI@MDU granted project, the collaboration between MDU, Gold Sentinel Inc., and the University of Waterloo in Canada. |
Start date: | |
End date: | |
Prerequisites: |
Proficiency in Python. Understanding synthetic data generation and simulation is highly valuable.
|
IDT supervisors: | Samaneh Zolfaghari |
Examiner: | Masoud Daneshtalab |
Comments: |
For further information, please contact samaneh.zolfaghari@mdu.se |
Company contact: |
Gold Sentinel Inc., Canada |