Principal Machine Learning Researcher @ Theator.
Before that I was a Post-Doctoral fellow in Laboratory for Active and Attentive Vision, York University , Toronto / Vector Institute.
email: amir dot rosenfeld (some-kind-of-pastry) gmail.com
News
2020:
Feb: our ICLR Paper featured in MIT news
2019:
Dec : One paper accepted to ICLR2020
April 29 : Invited talk at Princeton's PIXL Lunch
March: Follow up work for "Totally Looks Like"
2018:
September: Our work on Perceptual Similarity accepted to ACCV 2018
April: One workshop paper and one extended abstract in CVPR 2018; One paper accepted to ICPR 2018
Nov: The Elephant in the Room featured in NY times
February: Work accepted at VSS 2018
2017:
July : Winners of Visual Decathlon Challenge at CVPR 2017 Pascal In Detail workshop. Talk slides , using our recent method
Media
Totally Looks Like mentioned in Quanta Magazine
Our work on The Elephant in the Room featured in the news:
New York Times: Artificial Intelligence Hits the Barrier of Meaning
Quanta Magazine: Machine Learning Confronts the Elephant in the Room
The Register: AI image recognition systems can be tricked by copying and pasting random objects
jiqizhixin.com: 「房间里的大象」:让目标检测器一脸懵逼
Import AI: Fooling object recognition systems by adding more objects
Le Scienze: L’elefante nella stanza dell’apprendimento automatico
Habrahabr: Machine learning: clash with a room elephant
Publications
A Constructive Prediction of the Generalization Error Across Scales , Jonathan S. Rosenfeld, Amir Rosenfeld, Yonatan Belinkov, Nir Shavit (ICLR 2020)
Intriguing Properties of Randomly Weighted Networks: Generalizing while Learning Next to Nothing , Amir Rosenfeld, John K. Tsotsos (Computer & Robot Vision, 2019 (Oral))
High-Level Perceptual Similarity is Enabled by Learning Diverse Tasks, Amir Rosenfeld, Richard Zemel, John K. Tsotsos (arXiv)
The Elephant in the Room, Amir Rosenfeld, Richard Zemel, John K. Tsotsos (arXiv)
try it yourself, download images used in the paper (be sure to view the readme file)
Youtube demo: https://www.youtube.com/watch?v=qcm3lL4PCC4
Totally Looks Like - How Humans Compare, Compared to Machines, Amir Rosenfeld, Markus D. Solbach, John K. Tsotsos (ACCV 2018; project homepage);
also appeared at Vision Sciences Society, May 18-23, St. Pete Beach, Florida) poster
Priming Neural Networks Amir Rosenfeld, Mahdi Biparva & John K. Tsotsos (CVPRW 2018, Oral @ MBCC Workshop)
Challenging Images For Minds and Machines , Amir Rosenfeld, John K. Tsotsos (arXiv)
Bridging Cognitive Programs and Machine Learning, Amir Rosenfeld, John K. Tsotsos, arXiv)
2017
Incremental Learning Through Deep Adaptation, Amir Rosenfeld, John K. Tsotsos (to appear, TPAMI)
Action Classification via Concepts and Attributes, Amir Rosenfeld, Shimon Ullman (ICPR 2018)
2016
Visual Concept Recognition and Localization via Iterative Introspection Amir Rosenfeld, Shimon Ullman. (ACCV 2016) .
Hand-Object Interaction and Precise Localization in Transitive Action Recognition Amir Rosenfeld, Shimon Ullman. Oral, CRV 2016 FRA_db dataset
2015
Face-space Action Recognition by Face-Object Interactions Amir Rosenfeld, Shimon Ullman. arXiv
2011
Extracting Foreground Masks towards Object Recognition Amir Rosenfeld, Daphna Weinshall. ICCV 2011.
29/10/2012 - a much improved version 2.0 now is available here (~350Kb)
Results on PASCAL2011 trainval set for download here (matlab format,~5Mb)
Invited Talks
April 29, 2019: Talk at PIXL lunch, Princeton CS
April 2nd, 2018: Talk at Vision Seminar at CSAIL
March 18, 2018: Talk at Weizmann Institute / Faculty of Math & CS
Visual Decathlon Challenge@CVPR 2017 Pascal In Detail workshop Talk slides
BGU Vision Seminar / 2016, Nov 16
HUJI Vision Seminar / 2016, June 19
Professional Service:
Reviewer for:
Frontiers in Computer Science 2022
CVPR 2017-2020
ICCV 2017,2019, 2021
TPAMI 2017,2018
ECCV 2018, 2020
AAAI 2020
WACV 2019,2020
BMVC 2018, 2019
Physics and Engineering Panel for PSC CUNY 2018,Neural Processing Letters, 2018