Read more: Click here to read the original, longer version of this story.
AS SOON as it鈥檚 on my head, Google Glass picks up my pulse. Late for a meeting, I鈥檇 had to scramble up to the third floor of the MIT Media Lab for this demo, and my racing heart shows it. I huff and puff while Javier Hernandez, the project鈥檚 leader, explains how it works. As he talks, my heartbeat settles down before my eyes: 79 beats per minute, 73, 69.
Hernandez has taught Glass to measure vital signs like pulse and respiration using just the headset鈥檚 built-in gyroscope, accelerometer and camera. They pick up the subtle head motion caused by the beating of the heart, and tease out the rise and fall of breaths from other movements.
Advertisement
Together with contextual data captured by Glass, Hernandez鈥檚 algorithms aim to give the wearer a window into their emotional state. It鈥檚 just one aspect of a new system, called SenseGlass, that is designed to make us more aware of our mental well-being, and ultimately control it better.
The work represents a leap forward in the field of affective computing, which seeks to build technologies that measure, respond to and influence our emotions. Rosalind Picard, who works with Hernandez, says Google Glass could be a powerful tool for making computers more emotionally attuned. 鈥淚t鈥檚 always been a challenge to have a computer understand something about your stress but not make it worse in the moment,鈥� she says. 鈥淚t鈥檚 been hard to get it right because we haven鈥檛 been able to monitor it in real time in a comfortable way until now. What鈥檚 really groundbreaking is that my ordinary eyeglasses could have that capability.鈥�
聯It鈥檚 a challenge to have a computer understand your stress but not make it worse in the moment聰
Powerful as it is to see your heart rate floating before you in a head-up display, biological data alone is not sufficient to infer emotion 鈥� context is also vital. For that, the researchers use Glass鈥檚 camera to record what the wearer is seeing. The researchers can analyse the visual data using apps they are developing, such as Smile Catcher, built by Niaja Farve, also at MIT, which counts the number of smiles flashed in the wearer鈥檚 direction each day. Such visual, social clues are combined with the wearer鈥檚 biological information.
Other apps exist that attempt to use Google Glass to discern emotions. Researchers at the Fraunhofer Institute in Erlangen, Germany, have just released an app called Shore, which identifies the emotions being expressed on the faces of people around you, while the start-up firm Emotient of San Diego, California, wants to use similar technology to let retailers measure your feelings.
SenseGlass is different, says Hernandez, in that the team鈥檚 goal is to help users take an active role in modulating their own emotions. 鈥淪ay you want to avoid depression,鈥� he says. Low variation in heart rate in response to events that should make people excited or nervous has been associated with depression, so a user might benefit from looking back on the system鈥檚 recordings. 鈥淚t鈥檚 about helping you to reflect on your daily life in a meaningful way,鈥� Hernandez says.
Jim Rehg of the Georgia Institute of Technology in Atlanta, who has worked on the SenseGlass project, says those who have trouble recognising or expressing their emotions, such as people with autism, might benefit by being able to use the device鈥檚 measurements to communicate how they are feeling to others. 鈥淭ech like this might be used to let a care provider better understand what is going on with the person they鈥檙e caring for,鈥� he says.
Rehg co-authored a paper on BioGlass, the aspect of SenseGlass that monitors pulse and breathing. He will present the work at the MobiHealth conference in Athens, Greece, in November.
The current version of the system has limitations. The sensors can鈥檛 pick up on pulse or breathing rate while the wearer is running or dancing, for instance, as such large movements drown out subtle motions. But the team say later versions should overcome this problem.
鈥淲hat they鈥檙e putting together has the potential to address all the challenges that come with real-world emotional monitoring,鈥� says Julien Penders, who directs the wearable healthcare programme at Imec, a medical company in San Francisco. 鈥淚n a single platform they managed to get context and physiological data, and not only to sense but also to deliver feedback. That鈥檚 very unique.鈥�
This article appeared in print under the headline 鈥淒on鈥檛 worry, be Glassy鈥�
