Element AI (STL@CSAIL) Tech Talk and Research Showcase

SystemsThatLearn@CSAIL member Element AI will be providing technical insights into the research interests and projects being conducted across the company. The Element AI Research team will be presenting a few short technical talks followed by a poster reception highlighting their current research areas and projects. Research will be presented on topics including, explainability, learning theory, domain adaptation, adversarial learning, few shot learning and much more. This event will engage both MIT faculty and students on common research interests and collaborative opportunities.

October 23rd 4-7PM: 4th floor R&D Commons, MIT CSAIL
Tech Talk: 4-5PM
Poster Session: 5-7PM

Food and refreshments will be provided! :)
For more information and registration please visit- https://www.eventbrite.com/e/element-ai-tech-talk-and-research-showcase-tickets-50921428363

About Element AI:

Element AI is an AI solutions and products provider that advances cutting-edge AI research and turns it into scalable products that make businesses safer, stronger, and more agile. Co-founded in 2016 by Jean François Gagné and leading AI researcher Yoshua Bengio in the deep learning hub of Montreal, the company is focused on pioneering an AI-First world by turning the world’s most important AI research into transformative business applications. The company has a deep commitment to advanced research, with a renowned faculty fellow network, the largest privately-owned Canadian artificial intelligence R&D lab, as well as a growing network of leading research partners.


TECH TALK- 4-5PM, 4th floor R&D Commons MIT CSAIL

Chief Science Officer and Co-Founder, Dr. Nicolas Chapados will provide an overview on the research directions being investigated within the company and how research is conducted across the fundamental, applied and product teams at Element AI. Research scientist, Karolina Dziugaite from Element AI’s fundamental lab will speak about her recent work focused on constructing generalization bounds to help understand existing learning algorithms and propose new directions. Applied research scientist, Perouz Taslakian from Element AI’s AI Core team will present work and demos on the team’s recent progress on representation learning and vision.

Nicolas Chapados - Chief Science Officer and Co-Founder, Element AI

Karolina Dziugaite - Research Scientist, Element AI

Perouz Taslakian - Applied Research Scientist, Element AI



POSTER SESSION- 5-7PM, 4th floor R&D Commons MIT CSAIL

For presenter bios and poster abstracts please visit the eventbrite page: https://www.eventbrite.com/e/element-ai-tech-talk-and-research-showcase-tickets-50921428363


Poster Title: Unsupervised Domain Adaptation with Similarity Learning

Pedro O. Pinheiro - Research Scientist, Element AI

Poster Title: Functionality-preserving Adversarial Learning

Ousmane Dia - Applied Research Scientist, Element AI

Poster Title: Deep Prior

Alexandre Lacoste - Research Scientist, Element AI

Poster Title: TADAM: Task dependent adaptive metric for improved few-shot learning

Boris Oreshkin - Research Scientist, Element AI

Poster Title: Pix2Scene: Learning Implicit 3D Representations from Images

David Vazquez - Research Scientist, Element AI

Poster Title: Revisiting Generalization for Deep Learning

Karolina Dziugaite - Research Scientist, Element AI

Poster Title: W2GAN: Recovering an Optimal Transport Map with a GAN

Amjad Almahairi - Research Scientist, Element AI