Deep Learning based Controllable Animation

Deep learning could revolutionise character animation by enabling the automatic transfer of poses from observations like 2D images to 3D characters. This project aims to tackle critical gaps in pose understanding and transfer and interactive editing, enhancing the analysis, creation and manipulation of human/animations in digital media.

Character animation, the art of bringing virtual characters to life, is poised for transformation with the integration of advanced deep learning techniques. This project endeavors to automate the process by understanding the status from observations like 2D images and transferring those pose/style statuses to 3D characters. By leveraging cutting-edge machine learning models, the project aims to produce accurate estimation of human movements and simplify the typically labor-intensive and complex tasks associated with character rigging and animation. This approach not only promises improved scene and object understanding but also enhances the versatility and realism of the animated characters. These advancements could benefit industries that rely heavily on realistic 3D animations such as video games, animated films, and digital twins.

To achieve its objectives, the project will 1) develop a robust 3D estimation/reconstruction module to interpret and convert observations like 2D images into 3D pose information, 2) refine this technology to allow for interactive editing and fine control of the animations, 3) study both an in-house collected video dataset with human participants for healthcare purposes and public datasets.

Informal enquiries about the project should be directed to Dr Chen indicating your proposal and including your CV with qualification details (copies of publications, transcripts and certificates).

Supervisor Bio

Dr. Chen has a track record of publications in computer vision and machine learning with recognitions like ICME 2018 Best Paper Award. Prof. Han is an established scholar in computer vision with industry experience. They have extensive research collaboration with UK universities like Imperial College London and University of Cambridge. 

About the Department

99 percent of our research is rated in the highest two categories in the REF 2021, meaning it is classed as world-leading or internationally excellent. We are rated as 8th nationally for the quality of our research environment, showing that the Department of Computer Science is a vibrant and progressive place to undertake research.

Candidate Requirements

The candidate should ideally have a good first degree or a master degree in Computer Science, or a relevant subject; solid mathematical background and programming skills; preferably, prior experience with publications in computer vision, machine learning and deep learning. The English language requirements must also be met by the start of the PhD.

How to Apply

To apply for a PhD studentship, applications must be made directly to the University of Sheffield using the Postgraduate Online Application Form. Make sure you name Zhixiang Chen as your proposed supervisor(s).

Information on what documents are required and a link to the application form can be found here – https://www.sheffield.ac.uk/postgraduate/phd/apply/applying

The form has comprehensive instructions for you to follow, and pop-up help is available.

Your research proposal should:

 -be no longer than 4 A4 pages, include references

 -outline your reasons for applying for this studentship

 -explain how you would approach the research, including details of your skills and experience in the topic area