2-year research programs setup to help solve problems facing the visual effects industry; areas of focus include the application of machine learning to solve VFX light transport simulation problems, and learned resolution - independent rendering for feature films.
Creative studio Framestore and Bournemouth University are seeking two research fellows to help drive forward the future of visual effects. Joining the Faculty of Media and Communications’ Centre for Applied Creative Technologies (CfACTs) and gaining access to Framestore’s world-leading teams, tech and software, the selected candidates will embark on two-year research programs to help solve key problems facing the VFX industry.
The research themes are:
Applying Machine Learning to Solve VFX Light Transport Simulation Problems: This will see the successful candidate using machine learning to explore and solve the challenges surrounding the scattering and diffusion of light.
Learned Resolution, Independent Rendering for Feature Films: This fellowship will see the successful candidate using machine learning to better balance super high-resolution rendering while reducing computational footprint and cost.
“Framestore’s Technology & Research Team comprises a diverse melting pot of computer scientists, engineers and physicists who are always striving for innovative solutions to take the company’s work to the next level,” explained Framestore’s Global Head of Software VFX Manne Öhrström. “This is a group of gifted technologists wholly focused on the industry’s future, and the work they do impacts every aspect of Framestore’s business. The potential for using machine learning in areas like lighting and rendering is huge, and we can’t wait to welcome two new research fellows to the team - we’re sure that their work will prove absolutely invaluable."
Candidates will need a doctoral degree, a first-class mind, and a passion for innovation. Applications are open until August 22 - find out more here.
To hear more about Framestore’s Technology and Research work, Öhrström, Leo Hills and Theodore Jones will be discussing the intersection of machine learning and real-time when it comes to rigging photoreal digihumans at this year’s DigiPro conference. Find out more here.
Source: Framestore
