Enhance! Neural Supersampling is Here! 馃攷
CrossMind.ai logo
Previously, TecoGAN AI was able to produce high resolution images and video from low resolution inputs. However, the method is not perfect when working with real time situations, like in video games. However, with supersampling (splitting pixels into more pixels to create a higher detail image), this particular case can be resolved. "Neural Supersampling for Real-time Rendering" by Lei Xiao, Salah Nouri, Matt Chapman, Alexander Fix, Douglas Lanman, and Anton Kaplanyan: https://research.fb.com/blog/2020/07/introducing-neural-supersampling-for-real-time-rendering/ https://research.fb.com/publications/neural-supersampling-for-real-time-rendering/