Allez Go: AI Fencing Referee

Technology in fencing is generally an underdeveloped field and automated referees present potentially significant benefits to the sport. Automated referees will offer a more consistent call compared to a group of human referees with slightly different interpretations of the fencing rules.

The gradual increase in online fencing videos over the past decade has allowed for novel technical projects in fencing that rely heavily on data, such as artificial intelligence. This study resulted in a state-of-the-art lightweight Temporal Convolutional Network to referee fencing bouts and classify actions as either a touch for the fencer on the left or the fencer on the right. To address this problem, we developed a pose estimation and audio analysis approach to autonomously referee fencing bouts. Using a custom dataset of international level fencing from the last 7 years, including ~4000 unique clips, our model achieved an accuracy of 89.1%, a 20% increase over previous state-of-the-art models. This model leverages advancements in human pose estimation to extract the position of both fencers and avoids high computational loads typically associated with CNNs. Additionally, it uses a novel technique to solve the issue of blade contact, a key component of refereeing fencing that was generally unaddressed in previous works. Our novel solution uses audio to ‘listen’ for the sound of blade contact rather than attempting to identify it visually.

Explore More!

Project Webpage

Quantifying the True Trade Value of NBA Players: A Data Driven Approach

This study investigates the drivers of NBA player valuation by analyzing performance metrics, advanced statistics, and salaries from the 2022–2023 season, using regression models to assess how performance relates to trade value. The findings show that while team payroll strongly correlates with team performance, individual player salaries often diverge from statistical output, highlighting the influence of factors beyond on-court performance.

Ali Q.

Mentored by Rami Abi-Akl

Integrating Pixhawk and Realsense Camera for Rover Autonomy

This study presents the design and implementation of a low-cost autonomous go-kart that uses a Pixhawk-4 flight controller and ArduPilot to enable waypoint-based navigation and speed control without human intervention. It also demonstrates how integrating an Intel RealSense depth camera improves safety through MAVLink-based obstacle avoidance, particularly in environments with obstacles.