2020 Microsoft Research Ada Lovelace Fellows
University of Toronto
David is a PhD student in the Machine Learning Group at the University of Toronto supervised by Professor Sanja Fidler. He is also affiliated with the Vector Institute for Artificial Intelligence. His research is focused on learning efficient representations that transfer across multiple domains and help to overcome the need for massive manually-labeled datasets. To this end, he is also interested in developing new learning algorithms and neural architectures that lead to better generalization. He has co-authored more than 10 publications in top-tier conferences including PolygonRNN++, STEAL, GSCNN, and one of the first attempts to bridge the gap between real and synthetic images for computer vision tasks via domain randomization.
University of California – Irvine
In a modern digitally-enabled society, access to advanced technologies and the ability to use them for the greater good are becoming increasingly essential for ethical and human challenges. The population of people over the age of 65 is growing at an unprecedented rate, with the latest US Census report predicting that as many as 1 in 5 people will be over 65 in the next dozen years. Physical challenges will abound as this group ages. Jazette’s work examines how we might combat social isolation and therefore improve mental, emotional, and physical health in people with dementia (PWD) and their caregivers through the design of virtual support technologies. This research seeks to understand this design space more deeply, prototype potential innovative solutions, and empirically validate these approaches with an eye towards not only contributing to the science behind technologies for aging but also potential creating life changing products for this huge and growing user population.
Georgia Institute of Technology
Aditi is a PhD student in Algorithms, Combinatorics and Optimization at Georgia Institute of Technology, advised by Santosh Vempala. Her research interests include randomized algorithms, efficient sampling and its applications, graph algorithms, complexity theory, and combinatorics. Her research at Georgia Tech involves developing faster algorithms for high dimensional sampling and convex optimization. Progress on sampling algorithms has lead to many useful tools, both theoretical and practical and it forms an essential part of algorithms for optimization, integration, statistical inference, linear programming, approximate counting, and other applications. Aditi’s past work includes finding faster sampling methods for polytopes using ellipsoidal Markov chains.
Sarah A. Riley*
Sarah is a PhD student in Information Science at Cornell University. She studies municipal algorithmic systems and social inequality and is interested in novel methods for detecting, measuring, and correcting bias. Her current project explores these issues in the context of dataset shift and pretrial risk assessment systems.
Wenqi’s research interests lie at the intersection of computer vision and graphics. She is particularly interested in applying computer vision to enable creativity and augment human perception of reality. Her current research focuses on simplifying image and video editing by leveraging 3D geometric reasoning. In the future, she hopes to continue exploring new machine learning models to improve scene understanding and enable smarter image synthesis and manipulation.
Hiwot Tadese Kassa*
University of Michigan
Hiwot’s thesis focuses on designing hardware accelerators that are flexible in serving applications from different domains such as AI, machine learning, and graph analytics by targeting the algorithms that are at the core of the computation, rather than the fully-packaged application. These are done based on a language- and compiler-level framework that can identify computational patterns in an application and map them to the best-fitting hardware accelerator. The goal of the project is to help in the advancement of emerging algorithms and applications because it addresses the growing complexity and computation demands of these domains while alleviating the need for application developers to become experts in the wide range of hardware accelerators that will serve the field. Similarly, it will support the design of more effective hardware accelerators, by gathering data on the accelerator’s characteristics that are in highest demand by applications.
University of Washington
Technology has brought new opportunities for marginalized people to collaborate and form community. But it has also brought new forms of control, collapsing contexts, and demanding standardized and singular ways of representing human complexity. Os Keyes studies how this might be partially reversed – how we go about building technologies that enable plural ways of being, knowing, and doing, with a particular focus on trans identities and lives. Their past work has looked at the impact of facial recognition on trans populations: current and future research directions include how datasets represent gender and the ways in which scientists go about building these representations. They can be found at https://ironholds.org.
Lydia T. Liu*
University of California, Berkeley
Lydia T. Liu is a PhD student in Computer Science at the University of California, Berkeley, advised by Moritz Hardt and Michael I. Jordan. Her research aims to establish the theoretical foundations for machine learning algorithms to have reliable and robust performance, as well as positive long-term societal impact. This involves developing learning algorithms that have strong guarantees and analyzing their distributional effects in dynamic or interactive settings.
Embodied virtual avatars in immersive virtual reality (VR) can powerfully affect users’ perception, cognition, and behavior. Some effects can be detrimental to users without their knowledge. Divine’s research seeks to understand the role of implicit biases and embodied avatars. His goals are to determine when avatars in VR games, VR entertainment, and VR educational content are likely to produce undesirable changes in implicit racial bias. Subsequently, create guidelines resulting in a framework concerning how to design VR content to minimize these undesirable changes in implicit racial bias and other unwanted biases. It is critical that this research is conducted while the VR market is still developing and can benefit from innovation, feedback, and ideas surrounding social good.
University at Buffalo
Our skin and the microbes that inhabit it have evolved in unison. Interactions between them provide the first line of defense against invading pathogens and allergens. These interactions modulate an immune response, impacting autoimmune skin disorders like psoriasis. Izzy’s research uses an evolutionary framework and Bayesian network theory to understand the causal relationships within this complex and dynamic system. Based on real patient data, they apply modeling and simulation-based approaches to identify microbes that exacerbate immune responses in individuals with psoriasis. The results of their research will have immense impacts on the future of personalized healthcare.