Artificial Intelligence
Building intelligent machines that complement human reasoning and extend what people can do.
Publication
Statistical Adaptive Stochastic Gradient Methods
Project
Micro Co-design
The massive scale of cloud infrastructure services enables, and often necessitates, vertical co-design of the infrastructure stack by cloud providers. Micro co-design is a Minimally invasive, Cheap, and retro-fittable approach to co-design that extracts efficiency out of existing…
Group
Ability
The Ability team conducts research at the intersection of HCI and AI, with the aim of developing innovative technologies that extend the capabilities of and enhance quality of life for people with disabilities (long-term, temporary…
Video
Accessibility – How Microsoft is making technology accessible to more people
Researcher Meredith Ringel Morris discusses three projects highlighted at CHI 2018 which focus on improving technology for people with low-vision or blindness. Topics include featured paper Caption Crawler, Rich representations of visual content for screen…
Project
Accessible Information Retrieval
Web search is an important modern literacy skill, necessary for fully participating in many aspects of society including education, employment, commerce, and entertainment. However, web search is particularly challenging for people with print disabilities, such…
Project
AI Fairness and Disability
This project examines issues around AI FATE (Fairness, Accountability, Transparency, and Ethics) and Responsible AI, with a particular focus on how AI systems impact people with disabilities. This includes ensuring that mainstream AI tools are…
Project
Augmentative and Alternative Communication
ALS (amyotrophic lateral sclerosis) is a degenerative neuromuscular disease; people with late-stage ALS typically retain cognitive function, but lose the motor ability to speak, relying on gaze-controlled AAC (augmentative and alternative communication) devices for interpersonal…
Project
Resource-efficient ML for Edge and Endpoint IoT Devices
Our objective is to develop a library of efficient machine learning algorithms that can run on severely resource-constrained edge and endpoint IoT devices ranging from the Arduino to the Raspberry Pi.
Tool
Gaussian Process for Any Neural Architecture
Gaussian Process for Any Neural Architecture: Reference Implementations This repo is a companion to the paper linked below, which shows that the Gaussian process behavior arises in wide, randomly initialized, neural networks regardless of architecture.…