POST-DOCTORAL RESEARCHER – Deep Imaging Rendering
InterDigital Research, Rennes, France
InterDigital (https://www.interdigital.com/) develops mobile and video technologies that are at the core of devices, networks, and services worldwide. We solve many of the industry's most critical and complex technical challenges, inventing solutions for more efficient broadband networks, better video delivery, and richer multimedia experiences years ahead of market deployment. InterDigital has licenses and strategic relationships with many of the world's leading technology companies. Founded in 1972, InterDigital is listed on NASDAQ and is included in the S&P MidCap 400® index.
Our VFX research team provides exclusive R&D innovation to Technicolor, world-leading visual effects company, and creator of the pictures of the main blockbuster movies produced each year (such as The Lyon King, The Jungle Book…). This project is a collaboration with Technicolor and aims at improving their production pipeline.
Interdigital develops core AI based image and data processing related to real and visual effects (VFX) content creation. Leveraging research collaboration with Technicolor / MPC, Interdigital identifies content creation bottlenecks and develops dedicated solutions to be applied in real production workflows.
Ray tracing is involving Monte Carlo rendering as one of the most expensive process in the production workflow for animation and VFX. Tradeoffs between image quality and computing time are applied through the use of denoising solutions (https://openimagedenoise.github.io/gallery.html). In the field of deep learning, new approaches to denoise uncompleted renderings have been considered as the solution for the above tradeoff problem. Different architectures have been proposed, considering different input features provided by the renderer (see e.g.  ). Yet the temporal consistency of the denoised images remains an open problem, and fur/hair materials are notoriously challenging to handle. InterDigital proposes a post-doctoral position to investigate how convolutional neural networks can push further the state of the art in this area.
TASK & RESPONSABILITIES:
Review the literature and implement state-of-the-art solutions
Develop new solutions in line with VFX workflows use cases
Potentially integrate successful solutions within Technicolor production workflow
PhD in computer graphics with a strong background in rendering
Some background also in asset creation
Some background with RenderMan and Arnold
Experience in Deep Learning
Software development skills: python/TensorFlow
Communication skills: efficient exchanges to be engaged both with fellow researchers at InterDigital and content production teams
Fluent English mandatory. French would be a plus.
 T. Vogels, F. Rousselle, B. McWilliams, G. Röthlin, A. Harvill, D. Adler, M. Meyer, J. Novák, “Denoising with Kernel Prediction and Asymmetric Loss Functions”, ACM SIGGRAPH 2018
 K.-M. Wong, T.-T. Wong, “Robust deep residual denoising for Monte Carlo rendering”, SIGGRAPH Asia 2012