UC Berkeley gesture recognition system targets wearable devices

University researchers are working on a tiny chip that could be used for gesture recognition on smart watches and headsets. Unlike devices that use light to recognize hand gestures, this new approach relies upon sound waves.

The ultrasonic gesture recognizer technology, called Chirp, is being developed by research engineers at the University of California at Berkeley's Swarm Lab and Sensor and Actuator Center. Their research has progressed far enough for them to form a spinoff company, called Chirp Microsystems, to develop and market the technology, according to MIT Technology Review.

According to the Swarm Lab's website, current optical 3D imagers for gesture recognition, such as Microsoft (NASDAQ:MSFT) Kinect, suffer from large size and high power consumption. Further, their performance depends on ambient illumination, and they generally cannot operate in sunlight.

"These factors have prevented widespread adoption of gesture interfaces in energy- and volume-limited environments such as tablets and smartphones," the lab said.

According to the lab, using sound rather than light can result in "chip-scale solution size, low power consumption, and ambient light insensitivity."

The lab said its research was inspired by medical ultrasound and is specifically focused on building "an ultrasonic 3D range sensor system suitable for gesture-recognition using batch-fabricated micro-machined aluminum nitride (AlN) ultrasonic transducer arrays and custom CMOS electronics."

Chirp relies upon two computer chips, an ultrasound chip and another custom chip that sends and receives electrical signals from the ultrasound chip. Similar to radar, by generating sound wave pulses and then listening for their echoes to return, the Chirp system can calculate how far away something is. Swarm's system also includes a watch battery that can run for 30 hours thanks to the system's low power consumption.

Chirp technology promises to be small enough for implantation into wearable gadgets such as smart watches and head-mounted computers such as Google Glass (NASDAQ:GOOG), which don't have enough room to accommodate useful touch screens.

"There aren't a whole lot of options of what you can do on a touch screen when it's about the size of a quarter or so," Richard Przybyla, a graduate student at the UC Berkeley Sensor & Actuator Center, who designed the ultrasound chip, told MIT Technology Review.

Przybyla said one goal is to determine a basic set of gesture commands that can be programmed into Chirp-enabled devices. For example, a user could move their hand away from the smartphone screen in order to zoom in on a photo. Researcher may graduate from tracking hand motions with Chirp to individual finger tracking, he said.

For more:
- see this Swarm Lab webpage
- see this MIT Technology Review article
- see this phys.org article

Related articles:
CCS Insight: Despite innovation, challenges remain for smart wearable devices
Yankee Group: Taking wearable tech to the next level
In wearables, more questions than answers
ZTE eyes wearable computing market, backs Windows Phone
Which companies will play in the market? Is it time for smart watches?
Youi Labs gesture-based interface named 'best new idea' at CableLabs Innovation Showcase
Lowenstein's View: Smartphone ennui-- and what to do about it
Galaxy S4 could usher in more gesture-based apps

Suggested Articles

In 5G, AI must take center stage in how various functions and nodes in the network work in sync for the highest performance and lowest cost.

The C-Band Alliance (CBA) now says the U.S. could see billions of dollars going to the U.S. Treasury if its auction of C-band spectrum gets approved.

For the interoperability test, the companies integrated a bunch of their virtual network functions (VNFs).