OUR SPEAKERS

speaker_badge_banner_red.png
Share on:
Asset 14icon.png
Asset 39icon.png
Asset 12icon.png

Sima Sabah is a senior Computer Vision Researcher, combining deep learning techniques and classical computer vision expertise to solve challenging problems.

She holds a M.Sc. in Computer Science from Weizmann Institute, where she researched similarity in videos and discovered the wonders of singing acapella, and a B.Sc. in both Physics and Electrical Engineering from the Technion.

Her passion is teaching computers how to see and reason with the world, and she enjoys exploring all the awesome things AI enables us to do.

Sima Sabah

Computer Vision Researcher
Asset 12icon.png
Asset 1TWITTER.png
Asset 39icon.png
Asset 17icon.png
linkedin.png
twitter.png
facebook.png
github.png
English, Hebrew
Languages:
Asset 7TWITTER.png
Location:
Tel Aviv, Israel
Asset 7TWITTER.png
Can also give an online talk/webinar
Paid only. Contact speaker for pricing!

MY TALKS

Put Down the GANs - Flow-based Generative Models as an Alternative to GANs and VAEs

Data / AI / ML

Asset 12SLIDES.png
Asset 21talk.png
Asset 11SLIDES.png

Generative Adversarial Networks (GANs) are well known for their ability to generate realistic images, but if you ever tried to train them you know their adversarial nature makes them hard and unstable to optimize.

Flow-based generative models approach the image generation problem differently. One of their benefits is optimizing a single simple loss, making the optimization process easier. While variations of the method have been around for a while, many people are still not aware it exists. Given recent works that showcase the amazing capabilities of the method, you should get to know it too!

By the end of the talk you will understand the basic concepts behind Normalizing Flows for image generation, their unique characteristics, and how they compare to the popular GANs and VAEs methods.

Asset 1icon.png

Put Down the GANs - Flow-based Generative Models as an Alternative to GANs and VAEs

Completed

true

Visible

true

Order

3

Go to lecture page