Yani's portrait photo

Yani A. Ioannou

ORCID iD icon

Founder-in-Residence

Entrepreneur First

I'm currently a Founder-in-Residence at Entrepreneur First (EF), as a member of the first Toronto EF co-hort. Entrepreneur First is a pre-seed talent-based startup accelerator backed by the founders of LinkedIn, DeepMind and Paypal, as well as top VCs from Silicon Valley and Europe.

I was previously a Visiting Researcher (PostDoc) at Google Brain Toronto/Google AR Core, Sessional Lecturer at the University of Toronto, and a Research Scientist at Wayve.

I completed my PhD at the University of Cambridge in Oct. 2018, where I was supervised by Professor Roberto Cipolla, head of the Computer Vision and Robotics group in the Machine Intelligence Lab, and Dr. Antonio Criminisi. My PhD was supported by a Microsoft Research Ph.D. Scholarship and I collaborated with researchers at Microsoft Research Cambridge (UK) extensively.

I am currently interested in efficient deep learning, specifically sparse neural networks, and their application to problems in computer vision. I have in the past worked on 3D computer vision, towards methods for processing and recognizing objects in large point clouds.

Outside of research, I've worked on open source projects such as the Linux kernel and the Point Cloud Library (PCL).

Selected Publications

For a full publications list, see Google Scholar.

Paper coverpage

Gradient Flow in Sparse Neural Networks and How Lottery Tickets Win

Y. Ioannou, U. Evci, C. Keskin, Y. Dauphin

arXiv Pre-print
October, 2020

Paper coverpage

Rapid Classification of TESS Planet Candidates with Convolutional Neural Networks

H. Osborn, M. Ansdell, Y. Ioannou, M. Sasdelli, et al.

Astronomy & Astrophysics
Volume 633, Number A53
January, 2020

Paper coverpage

Scientific Domain Knowledge Improves Exoplanet Transit Classification with Deep Learning

M. Ansdell , Y. Ioannou , H. P. Osborn , M. Sasdelli, et al.

The Astrophysical Journal Letters
Volume 869, Number 1
December 5, 2018

PhD Thesis Coverpage

Structural Priors in Deep Neural Networks

Y. Ioannou

Ph.D. Thesis, Department of Engineering, University of Cambridge, Sept. 2017.
PDF (Print) BibTeX LaTeX Source



Paper coverpage

Deep Roots: Improving CNN Efficiency with Hierarchical Filter Groups

Y. Ioannou, D. Robertson, R. Cipolla, A. Criminisi

IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2017, Honolulu, Hawaii.
BibTeX Poster

Paper coverpage

Training CNNs with Low-Rank Filters for Efficient Image Classification

Y. Ioannou, D. Robertson, J. Shotton, R. Cipolla, A. Criminisi

International Conference on Learning Representations (ICLR) 2016, San Juan, Puerto Rico.
BibTeX Poster Models

Paper coverpage

Decision Forests, Convolutional Networks and the Models in-Between

Y. Ioannou, D. Robertson, D. Zikic, P. Kontschieder, J. Shotton, M. Brown, A. Criminisi

Microsoft Research Tech Report
MSR-TR-2015-58, April, 2015.
BibTeX

Paper coverpage

Emergency detection and response system and method

A. Mihailidis, Y. Ioannou, J. Boger, J. Gastle

United States Patent Application, 2013/0100268 A1, Apr. 25, 2013



Paper coverpage

Difference of Normals as a Multi-Scale Operator in Unorganized Point Clouds

Y. Ioannou, B. Taati, R. Harrap, M. Greenspan

3D Imaging, Modeling, Processing, Visualization and Transmission (3DIMPVT) 2012, Zürich, Switzerland. BibTeX Poster

Master's Thesis Coverpage

Automatic Urban Modelling using Mobile Urban LIDAR Data

Y. Ioannou

Thesis (M.Sc. Computing), Queen's University, Canada, 2010.

Teaching

Courses I've Taught Recently

Talks

Invited Talks and Lectures

McGill University, Montréal, Canada

Structural Priors in Deep Neural Networks

March 12, 2018
Slides (PDF)

University of Toronto, Toronto, Canada

Structural Priors in Deep Networks

August 29, 2017
Slides

KAIST, Daejeon, South Korea

Restricted Connectivity in Deep Neural Networks

April 17, 2017
Slides

Microsoft Research, Cambridge, UK

Restricted Connectivity in Deep Neural Networks

March 21, 2017

Curriculum Vitae/Resume

See LinkedIn for more information.

Contact Me

Preferably contact me about careers on , and research on . Alternatively, e-mail me.

You can also find out more about me through one of the many social links below.