Physics- Based Image De-shadowing Using Local Linear Model

Hi everyone, i'm happy to announce another event before the holiday season. in this event Tamir Einy from Applied materials will share the details behind their sotA deshadowing model.

Schedule:
18:30 - Meet, Greet, Eat pizza and drink beer
19:00 - Physics- Based Image De-shadowing Using Local Linear Model - Tamir Einy
20:00 - Mingling

Looking forward to seeing you all.

Abstract
Image deshadowing algorithms remove shadows from images. This requires both detecting
where the shadow is and, once detected, removing it from the image. This work focuses on
the shadow removal stage. We follow a common physical shadow formation model and
learn its parameters using a deep neural network. Our model consists of an existing network
for shadow detection, and a novel network for shadow removal. The shadow removal
network receives the predicted mask of the shadow region and shadow image, and predicts
six parameters per pixel. Remarkably, this straightforward network architecture, that is
considerably smaller compared to alternative methods, produces better results on standard
datasets.

https://techtime.co.il/2022/07/24/applied-materials-57/

Bio
Tamir Einy is a senior Algorithm Engineer at Applied Materials, working on image denoising.
Previously, Tamir worked as an Algorithm Engineer at Pixellot, Intel RealSense and Silentium.
He holds an M.Sc. in Electrical Engineering from Tel Aviv University and a B.Sc. in Electrical
Engineering from the Technion.