This project was based on a Machine Learning research paper which talks about converting YouTube videos of human movements into physics simulated 3D character animations.

​

I worked as a producer and game designer on this project. This was a very challenging project that put me out of my comfort zone. I found it very rewarding and challenging to communicate with all the different stakeholders in the project (our clients at Google and our faculty advisors) and making sure that all their requirements were effectively communicated to my team. It was also really challenging to make sure we were on track to hand-off deliverables and scope the project to fit our timeline while also  meeting client requirements.

​

Our main goal was to discover and document various game design applications for this technology and create a Machine Learning pipeline based on the research paper.

​

We successfully built and delivered a pipeline which converts videos into 3D character animations and inserts them into a interactive game-like experience in Unreal Engine.

 

Please visit the project website for a detailed look into our development process, the challenges we faced and to view detailed game design documentation.