Session 1.A

preview_player
Показать описание
Welcome.
#27: NLDNet++: A Physics Based Single Image Dehazing Network.
#61: Distributed Sky Imaging Radiometry and Tomography.
#79: Towards Learning-based Inverse Subsurface Scattering.
#86: Unveiling Optical Properties in Underwater Images.
Рекомендации по теме
Комментарии
Автор

Hi everyone - if you have further questions, please write your questions via comment and the authors will asynchronously respond. Feel free to also direct message any authors on Slack if they are available, but not all authors will be necessarily on Slack

surenjayasuriya
Автор

Question regarding #79. Is there any comment regarding how "good" the forward rendering process need to be? Is it that rendering even with less sample can still be fixed by more iteration of back projection? Or do we need to make sure the renderer is at almost highest sampling number possible to guarantee meaningful gradients.

ChiaYinTsai