ABOUT
Project Mimicus is an interdisciplinary team of students at the Silicon Valley campus of the Entertainment Technology Center. We are working with Google to use cutting edge Machine Learning research to convert videos of human movements into 3D character animations. Our goal is to document our learning process and design a game using this technology while exploring a variety of possible game design applications.
Our project is based on a machine learning research paper (Learning Acrobatics by Watching YouTube) which converts videos of human movements into animations. Our goal is to explore possibilities of using this technology in the game industry. We are coming up with a wide variety of game design applications for this technology, based on which we are creating a game that tries to makes the best use of this technology
Read the Research PaperWHY WE ARE DOING THIS
To explore possibilities of using this technology in the game industry and answer the question “In its current state, can we make a game using this technology?”
WHAT WE ARE BUILDING
We are exploring a wide range of possible game design applications and documenting our learning process while trying to build a game which uses this technology.
OUR APPROACH
We are putting together a machine learning pipeline which goes from 2D video to animations in Unreal. We will take this pipeline and create an ideal game loop for it.
Our Team
Haran Kim
Haran received two MFA degrees from Indiana University majoring in Printmaking and Digital Art. She is a digital artist who has a great passion for the technical and creative aspects of the entertainment industry and is seeking opprtunities as an artist.
Artist | Designer
ETC PagePortfolio
Eliza Zhang
Eliza is a software engineer who is interested in rendering and machine learning. Working at the
intersection of technology and art is exciting to her.
Machine Learning Programmer
ETC PageFaculty Advisors
Melanie Lam
Carl Rosendahl
Latest Posts
Contact
Email us: etc-sv-mimicus@lists.andrew.cmu.edu
