Google Maps for mobile ||
I work to put a visually stunning map, powerful local search tool, and
full-featured navigation system in your pocket.
|
SqueezeBlock: A Squeezable Interface for Mobile Devices ||
Touch is underused as a human-computer interface. Our SqueezeBlock prototype
was an electro-mechanical system to realize a virtual spring. It could
programmatically alter its squeezability, shape, and texture patterns to
convey information to its user in an eyes-free fashion.
|
Remote Control Identity Inference || Accelerometers and key
press sensors embedded in embedded devices such as television remote controls
can be used to distinguish household members based on the unique way each
person wields the remote. This personalization capability can be used, for
example, to enhance digital video recorders with show recommendations per
family-member instead of per device.
|
Low Power Face Recognition || Mobile computer vision brings
new possibilities to everyday experiences. For example, face
recognition and visual location detection enable easy authentication and simple reminders as you walk through the world.
This project was about making cutting-edge computer vision and perception
algorithms practical on mobile devices through
platform-integrated machine learning, low power networking, and
parallelization in the network cloud.
|
InertiaDot (YouTube vido) ||
The InertiaDot is a compact self-contained battery-operated 6 degree-of-freedom
Bluetooth-wireless Attitude and Heading Reference System (AHRS). It features a
2000 degree per second rotational response and fits in a 27x30x8mm package
weighing only 9.8 grams. AHRS computation is done on-board, based on a non-linear estimation
filter developed for the DIY
Drones UAV DevBoard from work of [Mahoney
et al. 2006].
|
Mobile Sensing Platform || To research mobile inferrence of
user behavior and context, the University of Washington and Intel Labs
Seattle designed a device called the Mobile Sensing Platform (MSP) for real-time processing of sensor traces, activity
inference, and conversation understanding. We built 100 MSPs to give
to collaborators in the research community.
|
Place Lab || Place Lab established the viability of the radio-beacon location technology that is a core capability in most modern smartphones and mobile devices. Read an overview
paper or watch me give a
research talk about Place Lab.
|
Reno: Location Sharing || Reno was a peer-to-peer, mobile,
location-enhanced messaging service forshadowing modern commercial services
like Dodgeball, Foursquare, Google Latitude, and Facebook check-ins. We
developed Reno to investigate techincal and social issues around disclosing an
individual's location including why, when, and what people want to share about
their location; the effectiveness automatic location disclosure; strategies
to achieve plausible deniability; and understanding how place and activity
are used to communicate plans, intentions and provide awareness.
|
SpotON || The SpotON ad hoc radio location hardware was
early work to determine the position of individual nodes in an ad hoc wireless
network. My SpotON signal strength measurement circuitry was used in the radio
design of the original Berkeley motes.
|
The Location Stack || My old PhD thesis work on design and sensor fusion for location-enhanced computing.
|
|
Technology