News
January 23, 2023:
Our recent work on uncertainty quantification with generative models was featured in MIT News: putting clear bounds on uncertainty
December 17, 2021:
"RAISR: Rapid and Accurate Image Super Resolution" was selected for the 2021 IEEE Signal Processing Society Best Paper Award.
March 2, 2021:
I have been awarded the 2020 Eric and Sheila Samson Prime Minister’s Prize for Global Innovation in Smart Mobility and Alternative Fuels for Transportation (the world's largest monetary award in these fields).
Official announcement (Hebrew), Technion news, EE department news, and CS department news.
October 26, 2020
I have received the 2021-2022 Career Advancement Chair (CAC), Technion.
October 25, 2020:
The Washington Post will use CQR to estimate outstanding votes for the 2020 U.S. presidential election:
"The predictive core of our model is a quantile regression. [...] We can estimate the 0.05th and 0.95th — for example — quantiles, which gives us a first estimate for our prediction intervals. Critically, this means that we did not make any distributional assumptions to get these prediction intervals. This is a big change from our Virginia model where we assumed turnout to be jointly normal!
Unfortunately, these prediction intervals aren’t guaranteed to be correct because our model may not be specified properly or because the counties that have delivered early results are particularly unrepresentative. To solve this problem, we use a method called conformal prediction. A modern statistical method that adjusts the prediction intervals generated by quantile regression to make them correct."
July 10, 2020:
Presented RED at SIAM Conference on Imaging Science (prize lecture).
December 21, 2019:
I have awarded the SIAG/IS Early Career Prize 2020.
December 10, 2019:
I have been awarded the Koret Postdoctoral Scholarship 2019–2020, Stanford University.
June 29, 2019:
Presented "deep knockoffs machines" at JSM (invited talk).
December 20, 2019:
Thanks to Ben Adcock, Simone Brugiapaglia, Yaniv Plan, and Ozgur Yilmaz for organizing a session on Sparse Recovery, Learning, and Neural Networks. It was a pleasure to attend and present our analysis of the stability of CNNs to adversarial attacks.
February 5, 2018:
Ofer Shacham, Engineering Manager for Pixel Visual Core:
"Like the main Pixel camera, Pixel Visual Core also runs RAISR, which means zoomed-in shots look sharper and more detailed than ever before."
Google Blog - Use Pixel 2 for better photos in Instagram, WhatsApp and Snapchat
October 24, 2017:
Together with Prof. Michael Elad, we constructed a Massive Open Online Course (MOOC) on Sparse Representations on the edX platform.
You are more than welcome to join us!
October 20, 2017:
RAISR is used in Pixel 2/XL Phones for digital zoom.
October 20, 2017:
RAISR is used in Google Clips for high quality image export.
April 06, 2017:
In our recent paper we analyzed fully-connected and convolutional neural networks (CNN) through the eyes of sparse representations. This work was described in the magazine of the Technion as"...unprecedented theoretical foundation to one of the hottest scientific fields today – deep learning."
January 11, 2017:
RAISR saves you bandwidth! According to Google Blog:
"...We’re already applying RAISR to more than 1 billion images per week, reducing these users’ total bandwidth by about a third. In the coming weeks we plan to roll this technology out more broadly — and we’re excited to see what further time and data savings we can offer..."
CNET - Google AI expands your photos to shrink your mobile data usage
PC Magazine - Google RAISR intelligently makes low-res images high quality
The Verge - Google is using machine learning to reduce the data needed for high-res images
ZDNet - Google sets RAISR to work: Hi-res images now hog less Google+ mobile bandwidth
Bit Tech - Google slashes image file sizes with RAISR
DPReview - Google brings RAISR smart image upsampling to Android devices
The Tech Report - Google RAISR upsamples thumbnails for massive bandwidth savings
Android Headlines - Google’s new RAISR image algorithm coming to Google+
Android Central - Google is bringing its bandwidth-saving RAISR image processing to your Android phone
Android Police - Photos on Google+ receive the machine learning treatment to help save on bandwidth usage
Gadgets 360 - Google+ uses machine learning to display high resolution images at a third of the bandwidth
December 15, 2016:
RAISR is being used in Motion Stills, as stated in Google Research Blog:
"...Last month, we published the details of our state-of-the-art RAISR technology, which employs machine learning to create super-resolution detail in images. This technology is now available in Motion Stills, automatically sharpening every video you export..."
November 14, 2016:
RAISR was described in Google Research Blog and featured in various websites, such as:
PetaPixel - Google’s RAISR uses machine learning to enhance low-res photos
Andoid Police - Google announces RAISR, a method of upscaling images with machine learning
The Verge - Google's prototype machine learning software lets you enhance low-res photos
New Atlas - Google's RAISR sharpens low-resolution images using machine learning
DPReview - Google RAISR uses machine learning for smarter upsampling
Android Authirity - Google reveals RAISR: an image enhancement tech which uses machine learning
The Tech Portal - Google debuts RAISR, a machine learning technique to upscale low-res images
Android Headlines - Google’s RAISR will upscale images using machine learning