HCII Seminar Series - Colin Lea
Speaker
Colin Lea
When
-
Where
Newell-Simon Hall 1305 - No public livestream this week
Description
This is the annual Sujeath Pareddy Memorial Lecture.
"Speech, Signing, & AI: Accessibility for All"
Speech-oriented interaction is ubiquitous in consumer technology, yet most systems are not fully accessible for the millions of people who have speech or communication disabilities. Whether due to differences in articulation, fluency patterns, or motor control, speech recognition models tend to perform poorly for these individuals. The first part of this talk will dive into our early studies aiming to better understand characteristics and needs of this user group, including the high interests and usefulness of these tools especially for people with severe motor-speech disabilities. We offer solutions ranging from shorter-term product features our team has already released, to progress towards longer-term research in collaboration with academic partners.
The second part of the talk shifts focus to speech generation alternatives, looking at our initial progress on sign language video generation, with a research goal of improving communication between hearing and non-hearing individuals. Throughout the talk, we emphasize the important of human-centered, accessibility-first design of machine learning systems, and how this is critical towards enabling technology that truly works for everyone.
Speaker's Bio
Colin is a research scientist and manager at Apple, where his team focuses on making interactive technologies more inclusive for people with disabilities. His work is at the intersection of machine learning, HCI, and accessibility, with an emphasis on speech and multi-modal systems. Prior to Apple, Colin completed his PhD in Computer Science at Johns Hopkins University in 2017 and was a research scientist at Facebook Reality Labs until moving to Apple in 2020.
Host
Jeff Bigham