When the still photo camera went on the market at the end of the 19th century, people declared it the end of personal privacy, a recent post on the Google Plus social-networking site noted. This insight was provided by the developers of Google Glass, the wearable computer that looks like eyeglasses.
- Facebook buying Google Glass rival for $2 billion
- Intel to invest $15m in Israeli startup's smartglasses
- 21st-century alchemy: There's gold in that roll-up screen
- Startup of the week / Finally, your smartphone can zoom
- Be'er Sheva wooing cybersecurity firms
- Loads of work experience but no job? Join the club
- Israeli system gives virtual sight to the blind
- Little guy who believes in privacy takes on the giants
- New star of 'Grey’s Anatomy': futuristic 3-D Israeli medical technology
- Look Ma, no hands! Google's new dream cars won't even have drivers
- Intel expands in Israel via Germany
Cameras were initially banned in public parks, at national monuments and on the beach, but now there are more cameras than ever. And those 19th-century fears resurfaced when the first cell phones came out, the Google people noted.
Three technology trends have been key for companies trying to create computer eyeglasses: the market penetration of mobile phones, the capacity to identify gestures and three-dimensional movement, and algorithms that make it easier to decipher still pictures and videos. And just as Israel has been a leader in cyber-security, it’s an increasingly important center for computer-vision technology.
Israeli companies in the field include Mobileye, which develops car collision avoidance technology; BriefCam, a video surgery equipment maker; and Cortica, which recently raised $20 million for its image and video analysis business.
Multinational high-tech firms in the sector are now busy in Israel as well. They include Apple, which bought 3-D sensing firm PrimeSense, and Intel, which acquired Omek Interactive, a gesture recognition and tracking technology company. Qualcomm, General Motors and Samsung are also in Israel developing new interfaces between people and machines.
TheMarker has convened a panel of Israeli leaders in the industry to explore the subject in greater depth: David Mendlovic, chief executive of Corephotonics, which develops technology for mobile devices; Gideon Shmuel, chief executive of EyeSight, a gesture and object recognition firm; Shaul Gelman, cofounder of RealView Imaging, which is developing a hologram device for medical uses; Chen Sagiv, the founder of SagivTech, which is in the image and signal processing business; and Eden Shochat, founding partner of the Aleph venture capital fund and co-founder of face.com, a face recognition company that was sold to Facebook in 2011.
“The spillover to industry from what is being done in the field in academia is very extensive,” Sagiv says, adding that these efforts are being buttressed by the work of defense firms such as Rafael Advanced Defense Systems and the development centers of companies such as Intel. They generate know-how that filters into the entire industry.
“This combination of major companies that are developing technology on a global level and in a strong academic community is a unique mix,” Sagiv says. “It’s turning Israel into a powerhouse in computer-vision technology. If you look at fields such as 3-D cameras and support software, you see that a major portion of the global leaders are Israeli.”
In addition to Israeli defense firms, the computer-vision industry has benefited from Israel’s contributions in printing technology, a sector where Israel was present an early stage, Shochat of the Aleph Fund notes. Even more, Israel’s strength in work combining optics, computerization, display and software pays off, he says.
Meanwhile, Gelman’s RealView Imaging is developing a hologram technology device that would give doctors three-dimensional images of patients’ internal organs from various angles. Although the device’s design has not yet been made public, a promotional film shows a doctor facing a hologram of a heart that he can rotate with his hand and various gestures.
The computer-vision interest in Israel on the part of multinational companies such as Apple, Intel and Qualcomm contributes to the local scene. But in some respects it makes life harder for Israeli firms.
Over the past two years, the multinationals have hired hundreds of Israelis in computer vision, making it harder for local firms to recruit staff, says Mendlovic of Corephotonics. And Shmuel of EyeSight acknowledges that the availability of good staff has hampered the growth of his business.
At a convention in Barcelona this year, Corephotonics unveiled a cell phone camera with two lenses. “Breakthroughs in hardware are rare,” says Mendlovic, the company’s chief executive. “PrimeSense did it, but it only happens once every few years.”
Mendlovic himself was one of the founders of Eyesquad, a company acquired in 2007 by San Jose, California-based Tessera Technologies for $24 million. The pictures available on cell phone cameras in low light are much improved, Mendlovic says.
“Companies such as Apple and Samsung each have a team of hundreds of people trying to get the quality produced by cell phone cameras to approach that of regular cameras,” he says.
“Beyond the existing capabilities of digital cameras, the future road map includes developments focused on 3-D capabilities; pictures that can provide information about depth, making it possible to control a device through gestures; an advanced capability to alter pictures; and features changing the focus after the picture is taken.”
As Shmuel puts it, “There is no hardware without software, and vice versa.” At the Barcelona tech conference, EyeSight unveiled a partnership agreement with OPPO, a Chinese company developing a smartphone using EyeSight’s technology, enabling the user to control the device through gestures alone.
“Computerized vision technology is undergoing huge changes at the moment, both inside and outside academia,” Shmuel says. He cites technology making it possible for anyone in a room to adjust a television’s volume with a hand gesture.
Just as good as people
A few weeks ago, the MIT Technology Review noted that a team of artificial intelligence researchers from Facebook and Tel Aviv University had developed facial identification software that can determine with 97.3% accuracy whether two photos are of the same person.
The journal says the level of accuracy demonstrated by the program, DeepFace, approaches the 97.5% accuracy that people achieve. One member of the Facebook team, Yaniv Taigman, is a founder of face.com, which was sold to Facebook in 2011. He’s a leading member of another Facebook team to figure out how the latest advances in artificial intelligence could benefit the social-networking giant.
On the question of the confidentiality of user information, Shochat of the Aleph Fund, who was a founder of face.com, says he and his colleagues learned that the more consumers receive value, the more they are willing to forgo privacy.
“How many times have you forgotten the name of a person standing in front of you?” he asks. “When I can get his name in real time, there is real value to that, which ultimately may justify a certain degree of infringement on privacy.”
Commenting on Google Glass, Google’s wearable computer glasses, Mendlovic, says the product shows that photos are not enough. “The real capability to become integrated into people’s lives also requires analysis and an understanding of the picture,” he says.
Google Glass is just the beginning, Shochat adds. Currently, when you use Google Glass, the information you receive appears in the corner of the device.
“It’s not fully mature technology,” he says. “Google Glass currently lacks the ability to map out every pixel that you see, a capability, for example, provided by Israeli company Lumus.”
The moment that each pixel is surveyed, you can add information on the glasses screen in a rational format.
“The significant change is its ‘always on’ feature. Glasses with full surveying capabilities could make it possible — for example, when I hear a song — to have the Shazam icon appear,” he says, referring to a mobile app that picks up music being played and identifies the song and the performer.
“It would make it possible to identify the song and buy it. Since Shazam’s business model is based on splitting the profits on the sale of songs from online stores, the company’s capacity to reach the customer would grow immeasurably.”