vIRal Vision Capstone Project
Description
June 2020-December 2020
Northeastern University Department of Electrical & Computer Engineering
- Collaborated with a team of five other ECE students to develop a novel solution regarding addressing and enforcing social distancing requirements through the use of wireless sensor networks, machine learning algorithms, and the Internet of Things (IoT).
- Goal: Create a system to track human movement in public spaces, identify choke points and high-traffic hotspots, and provide visual and auditory indicators that enforce social distancing.
Overall System Design
High-Level Software Architecture
As a member of the high-level software sub-team, my responsibilities included:
- Helping develop a suite of motion-prediction (predictor) algorithms that model how people intend to move through a given space based on kinematic, behavioral, and social criteria.
- Creating a multi-classifier machine learning model the accepts predictions from the predictor algorithms enabling the system to determine which algorithm is most accurate given the situational and environmental context.
Predictor Algorithm Suite
Core ML Model Design
Random Forest Modeling Outline
