New Paper: DynIBaR Neural Dynamic Image-Based Rendering

Zhengqi Li, Qianqian Wang, Forrester Cole, Richard Tucker, Noah Snavely of Cornell Tech and Google Research have released a new paper that addresses the challenges of moving subjects due to motion, camera paths, and video length.

The results are able to achieve photorealistic and high quality renderings despite the above challenges. As demonstrated in the linked video, there are several implications for live sport coverage as well as high action moments. Although it will more than likely be too soon, perhaps we shall see the X Games utilizing the technology to showcase the athlete's prowess; however, if anything has been learned within the last year, it is to not underestimate the pace of technology!

For anyone that would like to read the original paper, please view it here as a PDF.