AI Tech Week: Talent
Programme
16:30 Introduction & Welcome
16:35 Talk #1 Transformers for Vision at Scale:
16:50 Q&A
16:55 Talk #2: (Academic) Road to Amsterdam
17:10 Q&A
17:15 Talk #3 On the Need for Diverse AI Talent
17:30 Q&A
17:35 Talk #4 AI Research in Industry: Opportunities & challenges
17:50 Q&A
17:55 Closing
18:00 End!
Chair: Sara Magliance
Assistant Professor of Intelligent Data Engineering
Talk #1 by Mostafa DeghaniMostafa Dehghani is a Research Scientist at Google Brain, where he works on machine learning, in particular, attention based models from new architectures to scaling them up. Before Google, he was doing a PhD at the University of Amsterdam with a focus on improving the process of learning with imperfect supervision. He explored ideas around injecting inductive biases into algorithms, incorporating prior knowledge, and meta-learning the properties of the data using the data itself, in order to help learning algorithms to better learn from noisy or/and limited data.
Transformers for Vision at Scale:
Transformers have become the de-facto standard for sequence modeling in NLP, and in the past year, we are observing its applications to computer vision. In this talk, I’ll discuss the success of Transformers on large data regime in vision application and show that not only reliance on CNNs is not necessary and a pure transformer applied directly to sequences of image patches can perform very well in a wide range vision tasks, but also, compared to CNN-based models, Transformers are much more efficient in terms of computational cost when scaled up.
Talk #2 by Ilaria Tiddi
I am an Assistant Professor of Hybrid Intelligence in the Knowledge Representation and Reasoning group of the Vrije Universiteit Amsterdam. My research focuses on creating AI systems that generate complex narratives using a combination of machine learning and knowledge representation techniques. As part of my research activities, I am editor-in-chief of the CEUR-WS Editorial Board and part of the Knowledge Capture conference (K-CAP) Steering Committee, while at the VU I coordinate the Network Institute Academy Assistant projects and am part of the newly formed AI&Health research centre.
(Academic) Road to Amsterdam:
In this talk, I will give an overview of my research on hybrid intelligent systems for generating explanations using symbolic background knowledge. I will do so by showing the academic path I followed over the past years, showing how different methods and different disciplines have helped shaping my research line
Talk #3 by Sennay Ghebreab
Sennay Ghebreab is an Associate Professor of Socially-Intelligent AI at the University of Amsterdam, program director Master Information Studies, and Scientific Director of Civic-AI Lab, a research lab for civic-centered and community-minded AI technology.
On the need for diverse AI talent:
The lack of diversity of people in AI and the problem of biased algorithms have largely been dealt with separately. These, however, are two sides of the same coin: lack of diversity in the workplace is interwoven with biased algorithms. Government agencies, universities and technology companies recognize the problem of biased algorithms and the importance of diversity in AI. And yet, they are unable to sufficiently diversify their AI talent pool. In this short talk I highlight why?
Talk #4 by Babak Ehteshami Bejnordi
Babak Ehteshami Bejnordi is a research scientist and manager at Qualcomm AI Research. Babak’s primary research interests are conditional computation and efficient image and video analysis. Babak did his PhD on developing machine learning algorithms for breast cancer diagnosis at Radboud University in the Netherlands. Before joining Qualcomm, Babak was a visiting researcher at Harvard University.
AI Research in Industry: Opportunities & challenges:
In this talk, I will present an overview of the various research projects we are working on at Qualcomm AI Research in Amsterdam as well as the various kinds of research collaborations we have with academia. I will also discuss the challenges and opportunities for starting a career in ML and my perspective of transitioning from academia to industry.