Nirupam Roy
University of Maryland, MD·Computer Science·Early-career
Sound carries far more information than the words it contains: where a voice is coming from, how a hand moves across a table, whether someone is approaching from behind. Most computing devices ignore this spatial side of audio because capturing direction usually demands several microphones spread apart, which does not fit in an earbud or a battery-free sensor.
- Fields
Computer Science
Engineering
Materials Science
- Topics
acoustic sensing
ambient computing
wearable devices
- Grants
- 2 known grants, through Jun 2028
What they work on
- Enable sensing of large-scale gestures and movements near walls, floors and tabletops using coded signal projection through passive acoustic micro-structures and an adaptive intermittent computing architecture for battery-free operation.
- Detect micrometer-scale motion and micro-gestures with cross-spectrum acoustic sensing and localization, processed by a lightweight neural attention method that allocates resolution selectively.
- Add physical context, such as location and motion, to voice and speech interactions across indoor and outdoor settings so that commands are interpreted with awareness of the surroundings.
- Build a microstructure-assisted acoustic front end that recovers sound direction from a single microphone and encodes it into compact representations for a low-power on-device neural network.
- Align spatial acoustic features with speech embeddings in a language understanding model, and assess user requirements, market fit and technical constraints for wearable deployment through stakeholder interviews.
More about the research
The work here builds hardware and algorithms that recover rich spatial information from very small, very low-power acoustic systems, and then uses it to let everyday environments and wearables understand human activity.
Why it matters
Earbuds, hearables and voice assistants currently cannot tell where a sound came from, which limits their usefulness in busy real environments. Single-microphone spatial hearing could give users directional safety alerts, clearer separation of speakers in meetings, and better accessibility support. The ambient sensing side could allow existing buildings to be retrofitted for health and recovery tracking, assistance and security without requiring people to wear or carry devices.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
Phuc Nguyen
University of Massachusetts, MA·Computer Science·Early-career
Build computational models that translate jaw movement and faint facial vibrations, captured by an ear-worn sensor, into speech sounds and commands so users can control devices silently.
Mimi Xie
University of Texas at San Antonio, TX·Computer Science·Early-career
Redesign and optimize tiny machine learning models for battery-less, energy-harvesting IoT devices so that inference and decisions still happen on time under strict power, memory, and computation limits.
Yang Zhang
University of California-Los Angeles, CA·Computer Science
Identify the information and resource needs of novice and expert developers of self-sustaining, interaction-powered systems through observation of programming sessions, interviews and surveys.
The school
University of Maryland: 49 labs on PI Pages, $1.5B a year on research, 602 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.