Cartwheel.

Skeleton Animation x Cartwheel.

Currently working to build the best web based text-to-animation platform for veteran 3D artists and animation-curious beginners.

Python, Maya, Blender, GCS, AWS, Monocular computer vision system

Unity.

Facial Animation x Unity.

Working closely with our clients and machine learning scientists, I developed data preprocessing and post processing tools to best prepare the data and showcase the power of our machine learning tools. Throughout my time on the Ziva Dynamics team at Unity, we built pipelines for:

  • skeletal motion capture data captured using Vicon systems

  • facial deformation data captured using head mounted cameras

  • audio performances to drive our facial deformation model

I also spent time on set with clients to oversee and teach their technicians how to best capture and frame their actors’ performances for our model. For internal dataset building, I headed our internal data shoots to build out our training data based on the model’s physical and semantic needs.

Maya, FFMPEG, Vicon Motion Capture System,

Internship Project.

Internship x Ziva Dynamics.

In my role as a software engineering intern, I have had the privilege of working closely with VFX artists to create an Unreal Engine (UE4) demo. This project showcased the Ziva Real-Time muscle simulation technology. Throughout this project, I collaborated with my co-developer to implement a wide range of features, including multiple Non-Player Characters (NPCs), UI interfaces, camera controls, walking and flying AI controllers, behaviour trees, animations, montages, notifies, footprints, and Inverse Kinematics for a non-standard skeleton. Additionally, I utilized MotionBuilder, Vicon Shogun, and Maya, to record and create customized animations for the demo creature.

Unreal Engine, Maya, Motion Builder, Motion Capture, NPC Programming, Zbrush.

Capstone Project.

What’s Art Got to Do With It?

As AI technology becomes increasingly prevalent in the creative landscape, understanding its construction and limitations is paramount. Our focus is on creating a transparent AI model, trained on The Met dataset, to classify "good art" and "not art," shedding light on the subjectivity and potential shortcomings of AI in evaluating qualitative data. We aim to address the concerns surrounding technology's impact on artistic expression and the growing fear of job displacement within the commercial creative sector. Through this exploration, we seek to empower artists with a foundational understanding of AI, its creative applications, and the potential consequences of irresponsible AI usage.

Python, YOLO, Roboflow, googleColab, LabelImg

Website Development.

PetSitter

This pet-sitter website application prototype was designed to connect pet owners with reliable local pet sitters. Developed using JavaScript, React, OAuth, and MongoDB, our platform offered an instinctive user experience to create a profile, attach a pet to the owner’s profile, and search for local sitters in the area.

Javascript, React, Mongo DB, OAuth.

Virtual Reality.

Escape Room Series.

This collaborative project brought to life an immersive experience with three Virtual Reality escape rooms, each designed by a different group member. The project utilized Unity to craft each unique environment with a focus on puzzles and interactive elements, where players make their way through a single-player three-part maze.

Unity, C#.

Human-Computer Interaction.

Donation Mobile Prototype.

Taking the defence of Fairy Creek's old-growth forests as a case study, we delve into the challenges faced by decentralized civil disobedience movements. Activists require specialized support, including food, legal aid, medical assistance, media coverage, and more. Conventional communication tools have proven insufficient in overcoming logistical hurdles for sustaining decentralized movements. By reframing the problem and understanding user needs, we strive to create a platform that accommodates varying forms and amounts of donations. Our focus is on simplicity, accessibility, and user-friendliness, offering verified organizations and accurate information to support diverse user requirements.

Figma

Foundations of Artificial Intelligence.

Snake Game Pathfinding.

In this classic python-based snake game project, players control the snake's movement using hand gestures, adding an interactive twist to the traditional gameplay. Additionally, players compete against the computer on three levels of difficulty. Each level is represented by a different algorithm - level one features Depth-First Search (DFS), level two challenges players with Breadth-First Search (BFS), and level three tests their skills with the A* algorithm.