A Semi-trailer Truck Blind Spot Alert System for Detecting Cyclists with Transfer Learning


Charles Tang, Nicholas Medeiros (Advisor)

This study aims to reduce the number of truck-cyclist collisions, which are often caused by semi-trailer trucks making right-hook turns and poor driver attention to blind spots. To achieve this, we designed a visual-based blind spot warning system that can detect cyclists using deep learning and model environment tests. Achieved State-Of-The-Art accuracy of 0.95 AP (IoU: 0.5) on the Cyclist Detection benchmark.

Competing in the WRSEF and JSHS competitions.

[arXiv] [Poster] [Dataset]


Why One-Way Hallways Are Safer Than Two-Way Hallways


Charles Tang, Archanaa S Krishnan PhD (VT), Dr. Patrick Schaumont (WPI)

A simulation-based study was designed to evaluate different pedestrian hallway scenarios to minimize the risk of viral transmission. Our model tests 792 different pedestrian scenarios, including one-way hallways, two-way hallways, and rooms,to determine the risk in each scenario. Our results describe the hallway scenario that can minimize the risk of contact, and compare and contrast various hallway models and key parameters that influence risk.

I presented this research at the WPI 2021 Summer Research Showcase Conference.


Recent Coursework

MA2051: Ordinary Differential Equations

CS2102: Object-Oriented Design Concepts

MA2631: Probability Theory

DS1010: Data Science I

MA 2071: Matrices and Linear Algebra I

MA 1971: A Bridge to Higher Mathematics

AP Calculus BC

AP Physics C Mechanics

AP Computer Science A

AP Economics