Antoine Dangeard

Service technician for The Cube's 3D printing service: involves evaluating print viability, repairing printers, conversing with clients, executing prints, and logging orders.

Earned through mentorship in MECH 360 (Principles of Manufacturing). As McGill describes, the award applies to "quality students from a very recent offering of the class or laboratory who have a marked interest in teaching/learning and educational improvement".

An RNN-powered melody and drums generation website, built in the context of the 24-hour McGill AI Society Hackathon. Assembled and cleaned datasets, wrote the code for training models and generating music and drums, build the back-end and helped with front-end integration. Powered by Google's Magenta models, Flask and Javascript.

Competed with McGill Robotics' AUV in RoboSub 2023, and made it to the semi-finalist pool. I joined the team in September 2022 and built the computer vision, propulsion, and planning (decision-making) packages which were used at competition. I also helped with the implementation of a quaternion-based orientation PID, simulation, and state estimation.

Participated in dodge-ball intramural tournament at McGill. Reached semi-finals before elimination.

Won the first place prize for the advanced "Vision" challenge at RoboHacks 2023. The 24-hour challenge involved building a water-bottle based underwater robot, controlling it's buoyancy using a motor and syringe, building ailerons to enable forward movement, and using a camera and arduino computer to produce a 2D visualization of colored regions on the bottom of the pool.

Used object detection and color theory/fashion trends to evaluate the objective "quality" of a person's outfit from an image, then gave customized feedback to the user. Built in 48 hours using a YOLO object detection model to identify individual articles of clothing, OpenCV to extract the features of each item of clothing, and JavaScript and Flask for the website.

Achieved Top 10 lowest mean squared errors for an open data challenge aiming to predict university acceptance from data points like SAT scores, extra-curriculars, etc. Set up and trained a Ridge regression model with stochastic gradient descent in 48 hours (I unfortunately heard about the challenge just before the deadline).

Participated in robotics competition for middle and high schoolers. Helped code and build the Lego Mindstorm-based robot. Our team won 2^nd place in the 'WAYS' challenge and 3^rd in the 'TRANSPORT' challenge.