Use of cookies

We use our own and third-party cookies to improve our services. You can get more information and set your preferences.
Information on our cookies policy

Reject Cookies
Accept Cookies

Campus in maintenance
User and / or password incorrect
You have no active environment
Your access has been restricted. Consultation with the department of administration
Due to technical problems, the virtual campus is inaccessible. We are working to solve it. Sorry for the inconvenience.
Start   >  Events  >  Open Class: "Deep Learning for Graphs and Sets"
Date: 26-03-2019
Timetable: 6:30 pm to 7:30 pm
Location: Tech Talent Center (Badajoz, 73. Barcelona)
This event is full

Next Tuesday 26th March at 6:30pm, the UPC School Talent Tech Centre will host the Open Class: "Deep Learning for Graphs and Sets", by Adriana Romero.

Deep learning has achieved remarkable results in fields such as computer vision, speech recognition and natural language processing. More recently, there has been a surge in research on graph and set representation learning, leading to new state-of-the-art in different domains such as 3D vision, biological networks and social networks analysis.

In this talk, Adriana will present their work on Graph Attention Networks (GATs). She will start by reviewing early approaches to leverage neural networks for processing graph structured data, with special emphasis on graph convolutions, highlighting potential issues and motivating our work. Then, she will introduce GATs, a neural network architecture that leverages masked self-attentional layers to address the shortcomings of prior methods based on graph convolutions or their approximations. The second part of the talk will be devoted to set prediction and potential downstream applications. In particular, she will discuss the importance of modeling object co-occurrences in different set prediction tasks and datasets and she will showcase the problem of generating recipes from food images, where ingredients are predicted as sets.

The speaker

Adriana Romero is a research scientist at Facebook AI Research and an adjunct professor at McGill University. She is currently involved in a number of projects at the intersection of conditional generation, multi-modality and graph-structured data. Previously, she was a post-doctoral researcher at Montreal Institute for Learning Algorithms, advised by Prof. Yoshua Bengio. Her postdoctoral research revolved around deep learning techniques to tackle biomedical challenges, such as the ones posed by imaging multi-modality, high dimensional data and graph structures. Adriana received her Ph.D. from the University of Barcelona in 2015 with a thesis on assisting the training of deep neural networks with applications to computer vision, advised by Dr. Carlo Gatta. Her PhD included contributions in the fields of representation learning and model compression, with applications to image classification, image segmentation and remote sensing.

Registration is free but required.