Yedid Hoshen

Web http://www.cs.huji.ac.il/~yedid/   Email: yedid.hoshen@mail.huji.ac.il

Education

The Hebrew University of Jerusalem, Jerusalem, Israel

PhD in Computer Science, Sep 2012 - April 2016

• Supervisor: Prof. Shmuel Peleg
• Main interests: Deep Learning, Computer Vision, Speech Recognition.

Worcester College, University of Oxford, Oxford, UK

Masters in Physics, 2003 – 2007, First Class Honours

• Enrolled at Oxford at the age of 15 and obtained First Class results every year.
• Made a College Scholar for performance in Prelims.

Work Experience

Hebrew University of Jerusalem, Jerusalem, Israel 

Associate Professor, Nov 2023 - Present

Assistant Professor (Senior Lecturer), April 2019 - Nov 2023

Google, Tel Aviv, Israel 

Visiting Faculty Researcher, Nov 2022 - Nov 2024

• Research Focus: Image and video editing

Facebook AI Research, Tel Aviv, Israel 

Research Scientist, June 2018 - May 2019

• Research Focus: Generative models, source separation. 

Facebook AI Research, New York, NY, USA 

Postdoctoral Researcher, Jan 2017 - June 2018

Research Highlights:

• Efficient multi-agent predictive models (NIPS’17)
• GANs for new domains:

• Unsupervised exact image retrieval (ICLR’18)
• Unsupervised image to text matching (CVPR’18)  

• NAM - Replacing GAN methods by better non-adversarial alternatives (ECCV’18)
• Non-adversarial method for unsupervised word translation (EMNLP’18)

Taboola, Tel Aviv, Israel

Deep Learning Consultant, May 2016 - Dec 2016

• Leading the adoption of deep learning technology in Taboola’s product.

FST Biometrics, Rishon L’Zion, Israel

Deep Learning Consultant, July 2015 - Jan 2016

• Shaped the vision for the next generation of the product.
• Lead the company’s embracement of the Deep Learning revolution.

Google Research, New York City, NY, USA

Software Engineering Summer Intern, Summer 2014

• Trained Deep Neural Networks for Automatic Speech Recognition.
• Result: End-to-end acoustic modeling for raw multi-channel speech data.

Google [x], Mountain View, CA, USA

Software Engineering Summer Intern, Summer 2013

• Worked on computer vision algorithms for Google Glass -  GCam.
• Successfully completed two projects, helping shape key product decisions.

Israeli Space Programme, Israel

Satellite Mission Planning Algorithms Engineer, Nov 2009 - Sep 2012

• Devised innovative algorithms for satellite mission planning.

KBC Financial Products, London, UK

Equity Derivatives Trader, Associate Director, Sept 2007 - June 2009

• Successfully traded Proprietary Volatility Arbitrage strategies.

.

Press Reaction

The Verge, "Your wearable camera wobble is as unique as a fingerprint"

MIT Technology Review, "The Coming Era of Egocentric Video Analysis"

New Scientist, "Your telltale video camera shake can identify you".

The Boston Globe, "How your GoPro videos could give away your identity"

Publications

An Image is Worth More Than a Thousand Words:Towards Disentanglement in the Wild

Aviv Gabbay, Niv Cohen, Yedid Hoshen

Neural Information Processing Systems (NeurIPS 21’)

Scaling-up Disentanglement for Image Translation

Aviv Gabbay, Yedid Hoshen

International Conference on Computer Vision (ICCV 21’)

Deep Single Image Manipulation

Yael Vinker*, Eliahu Horwitz*, Nir Zabari, Yedid Hoshen

International Conference on Computer Vision (ICCV 21’) - Oral Presentation

Reconstruction-Based Membership Inference Attacks are Easier on Difficult Problems

Avital Shafran, Shmuel Peleg, Yedid Hoshen

International Conference on Computer Vision (ICCV 21’)

PANDA--Adapting Pretrained Features for Anomaly Detection

Tal Reiss*, Niv Cohen*, Liron Bergman, Yedid Hoshen

IEEE Conference on Computer Vision and Pattern Recognition (CVPR 21’)

Crypto-Oriented Neural Architecture Design

Avital Shafran, Gil Segev, Shmuel Peleg, Yedid Hoshen

International Conference on Acoustics, Speech and Signal Processing (ICASSP 21’)

Classification-Based Anomaly Detection for General Data

Liron Bergman, Yedid Hoshen

International Conference on Learning Representations (ICLR 20’)

Demystifying Inter-Class Disentanglement

Aviv Gabbay, Yedid Hoshen

International Conference on Learning Representations (ICLR 20’)

Non-Adversarial Image Synthesis with Generative Latent Nearest Neighbors

Yedid Hoshen, Ke Li, Jitendra Malik

IEEE Conference on Computer Vision and Pattern Recognition (CVPR 19’)

Neural separation of observed and unobserved distributions

Tavi Halperin, Ariel Ephrat, Yedid Hoshen

International Conference on Machine Learning (ICML 19’)

Towards Unsupervised Single-Channel Blind Source Separation using Adversarial Pair Unmix-and-Remix

Yedid Hoshen

International Conference on Acoustics, Speech and Signal Processing (ICASSP 19’)

Non-Adversarial Mapping with VAEs

Yedid Hoshen

Neural Information Processing Systems (NIPS 18’)

Non-Adversarial Unsupervised Word Translation

Yedid Hoshen and Lior Wolf

Conference on Empirical Methods in Natural Language Processing (EMNLP 18’)

NAM: Non-Adversarial Unsupervised Domain Mapping

Yedid Hoshen and Lior Wolf

European Conference on Computer Vision (ECCV 18’)

Unsupervised Correlation Analysis

Yedid Hoshen and Lior Wolf

IEEE Conference on Computer Vision and Pattern Recognition (CVPR 18’)

NAM - Unsupervised Cross-Domain Image Mapping without Cycles or GANs

Yedid Hoshen and Lior Wolf

International Conference on Learning Representations Workshop (ICLRW 18’)

Identifying Analogies Across Domains

Yedid Hoshen and Lior Wolf

International Conference on Learning Representations (ICLR 18’)

VAIN: Attentional Multi-agent Predictive Modeling

Yedid Hoshen

Neural Information Processing Systems (NIPS 17’)

Power to peep-all: Inference Attacks by Malicious Batteries on Mobile Devices

Pavel Lifshits, Roni Forte, Yedid Hoshen, Matt Halpern, Manuel Philipose, Mohit Tiwari, Mark Silberstein

Proceedings on Privacy Enhancing Technologies, (PETS 18’)

An Egocentric Look at Video Photographer Recognition

Yedid Hoshen and Shmuel Peleg

IEEE Conference on Computer Vision and Pattern Recognition (CVPR 16’)

Visual Learning of Arithmetic Operations

Yedid Hoshen and Shmuel Peleg

Thirtieth AAAI Conference on Artificial Intelligence (AAAI 16’)

Live Video Synopsis for Multiple Cameras

Yedid Hoshen and Shmuel Peleg

International Conference on Image Processing (ICIP 15’)

Speech Acoustic Modelling for ASR from Multichannel Waveforms

Yedid Hoshen, Ron Weiss and Kevin W Wilson

40th International Conference on Acoustics, Speech and Signal Processing (ICASSP 15’)

The Information in Temporal Histograms

Yedid Hoshen and Shmuel Peleg

IEEE Winter Conference on Applications of Computer Vision (WACV 15’)

Wisdom of the Crowd in Egocentric Video Curation

Yedid Hoshen, Gil Ben-Artzi, Shmuel Peleg

3rd Workshop on Egocentric (First-person) Vision. In conjunction with CVPR (CVPRW 14’)

Efficient Representation of Distributions for Background Subtraction

Yedid Hoshen, Chetan Arora, Yair Poleg and Shmuel Peleg

IEEE International Conference on Advanced Video and Signal-Based Surveillance (AVSS 13’)