Altek Brings AI To The Edge With Vision AI Chips, AI Commercial Surveillance Cameras, And 3D Depth Image Sensing Modules

27 Feb 2019

Download PDF version

Altek Corp., global digital imaging company based in Taiwan, is introducing three new technologies: Vision AI chips, AI commercial surveillance cameras, and 3D depth image sensing modules with the theme of ‘Bringing Vision AI to The Edge’. The company’s vision AI chip and AI commercial surveillance cameras enable high-resolution smart surveillance by supporting AI human/object detection, event detection and behavior recognition. The entire AI process is conducted in the camera and corresponding alerts are sent out according to the detection results. It reduces the network bandwidth for data transmission to the cloud and improves the AI detection speed. This technology can be utilized in multiple use cases, including in retail, restaurants and hospitals.

Altek also has collaborated with global tech leaders in providing commercial/home use surveillance system with edge vision AI capability"

“Being dedicated to more than 20 years in digital imaging, Altek believes that edge vision AI will unleash remarkable market demands.” Alex Hsia, Chairman of Altek Corporation said. “Recently, Altek also has collaborated with global tech leaders in providing commercial/home use surveillance system with edge vision AI capability, aiming to bring the surveillance environment in retail, manufacturing, factory and home to new level of experience.”

Altek – CyberLink Software Integration

In partnership with CyberLink Corp. (5203.TW), Altek’s 3D sensing solution has been integrated with CyberLink’s AI facial recognition engine FaceMe. The new integration strives to enhance the anti-spoofing capabilities of smart security systems, enabling them to accurately detect facial images and videos apart from live human faces.

“Anti-spoofing is vital for smart security and other AIoT applications. CyberLink’s FaceMe engine is capable of accurate anti-spoofing by utilizing an array of inputs including 2D & 3D cameras, human head motions, voice, and mainstream 3D imaging technologies including Altek’s active stereo technology. This allows FaceMe to achieve highly accurate human face pairing with faultless anti-spoofing results,” said Dr. Jau Huang, CEO of CyberLink.

“Through our collaboration with Altek Corp., we hope to implement FaceMe’s high-accuracy facial recognition technology into 3D camera solutions to deliver safer and more reliable facial recognition capabilities to hardware manufacturers of mobile phones, door access control systems, and a number of other applications.”

Edge Vision AI Technology

With Microsoft and Qualcomm, Altek also co-developed a Visions AI Developer Kit

Altek has also co-developed products with other global tech leaders to demonstrate its commitment to commercializing edge vision AI technology. With Qualcomm, the company has co-developed a commercial AI surveillance IP camera. This camera is powered by QCS603 platform with distributed edge AI computing power to enable AI detection of humans and objects within the recognition area in the camera. It also supports Amazon KVS so that users can leverage the AWS cloud platform to access, process and store streaming video with ease for further analysis by deep learning.

Visions AI Developer Kit

With Microsoft and Qualcomm, Altek also co-developed a Visions AI Developer Kit. By supporting Microsoft Azure IoT Edge and using Qualcomm AI Engine, it strengthens Azure’s deep learning service and delivers ultra-high resolution video quality, creating optimal capability for processing vision AI at the edge for various markets and applications.

Having established itself in the market of edge vision AI, Altek has integrated its software/hardware systems, chips and algorithms, and continues to collaborate with leading technology providers to leverage technological advantages and industry experiences to expand sales channels and provide omnidirectional services in the field of edge vision AI.

Download PDF version

People mentioned in this article

Dr. Jau Huang

Related categories
Access control software
IP cameras

In case you missed it

San Francisco Prohibits Deployment Of ‘Secret Surveillance’ Technologies

The mindset behind a new law to prohibit the use of facial recognition and other security-related technologies by San Francisco police and other city agencies is obvious in the name of the new ordinance: “Stop Secret Surveillance.” Ordinance To Stop Secret Surveillance The San Francisco Board of Supervisors passed the ordinance 8-1 with two abstentions on May 14, and there will be another vote next week before it becomes law. We have an outsize responsibility to regulate the excesses of technology precisely because they are headquartered here" The irony of such a law emanating from northern California, where tech giants promote the use of numerous technologies that arguably infringe on privacy, is not lost on Aaron Peskin, the city supervisor who sponsored the bill. “We have an outsize responsibility to regulate the excesses of technology precisely because they are headquartered here,” he told the New York Times. Regulating Facial Recognition Technology Although the facial recognition aspects of the ordinance have been the most publicized, it also targets a long list of other products and systems. According to the ordinance, "Surveillance Technology" means “any software, electronic device, system utilizing an electronic device, or similar device used, designed, or primarily intended to collect, retain, process, or share audio, electronic, visual, location, thermal, biometric, olfactory or similar information specifically associated with, or capable of being associated with, any individual or group.” Broadly interpreted, that’s a lot of devices. Includes Biometrics, RFID Scanners The ordinance lists some examples such as automatic license plate readers, gunshot detection hardware and services, video and audio monitoring and/or recording equipment, mobile DNA capture technology, radio-frequency ID (RFID) scanners, and biometric software or technology including facial, voice, iris, and gait-recognition software and databases. Among the exceptions listed in the ordinance are physical access control systems, employee identification management systems, and other physical control systems; and police interview rooms, holding cells, and internal security audio/video recording systems. The ordinance ban applies to city departments and agencies, not to the general public and exceptions include physical access control systems, employee identification management systems, and internal security audio/video recording systems Airport Security Not Part Of Ordinance The ban only applies to city departments and agencies, not to private businesses or the general public. Therefore, San Franciscans can continue to use facial recognition technology every day when they unlock their smart phones. And technologies such as facial recognition currently used at the San Francisco airport and ports are not impacted because they are under federal jurisdiction. Furthermore, the San Francisco police department does not currently use facial recognition anyway, although it has been deployed in places such as Las Vegas, Orlando, Boston and New York City. Safeguarding Privacy Of Citizens The ordinance appears to have a goal of avoiding government uses of technologies that can invade individual privacy, seeking to avoid worst-case scenarios such as an existing system in China that uses millions of surveillance cameras to keep close tabs on the Uyghurs, a Muslim minority population. Any new plans to use surveillance technology must be approved by the city government, and any existing uses must be reported and justified by submitting a Surveillance Technology Policy ordinance for approval by the Board of Supervisors within 180 days. Surveillance Technology Policy Banning use of facial recognition just when its capability is being realized is counterproductive But might such a ban on technology uses undermine their potential value as crime-fighting tools just when they are poised to become more valuable than ever? Ed Davis, a former Boston police commissioner, told the New York Times it is “premature to be banning things.” He notes: “This technology is still developing, and as it improves, this could be the answer to a lot of problems we have about securing our communities.” Technology development doesn’t happen in a vacuum and banning uses of facial recognition and other technologies just when their capabilities are being realized is counterproductive. We should be thoughtful, deliberate and transparent in how we embrace new technologies. However, discarding them out-of-hand using emotionally charged words such as “secret surveillance” does not promote the best use of technology to the benefit of everyone.

Considering Corporate Culture When Embracing Security Robots

Deploying security robots at a company is about more than providing and programming the hardware. There is also an element of “change management” involved in smoothing the way for robots to play a security role working side-by-side with human counterparts. Rising Popularity Of Security Robots As security robots increase in popularity, more companies are adapting to such cultural challenges "As security robots increase in popularity, more companies are adapting to such cultural challenges. Many Fortune 100, technology, finance and defense companies have begun using security robots, and some are asking to expand their implementation", says Travis Deyle, CEO and co-founder of Cobalt Robotics. "It is a complex solution that involves merging technology with people." “More people are looking at how they can deploy and test this technology, dipping their toes in the water,” says Deyle. “Financially the risk is low, but culturally it is pretty acute. It is a very visible piece of technology moving through your most sensitive spaces and interacting with employees.” Change Management “Doing change management and addressing the cultural implications inside the company are the biggest challenges we face,” Deyle adds. “We have to make sure that people know what the robot is there for, what it does and how it helps them. There is a social contract between companies and employees about what information is being collected and how is it being used.” The technical onboarding of a robot is the easy part, says Deyle. “The robot goes in, maps out the space; it takes about an hour. The bigger part is the cultural onboarding.” The process involves working closely with the company’s communications team to manage how the use of robots is messaged throughout the company. Deyle suggests doing a Q&A event where employees can touch and feel the robot and get comfortable. “We tailor the interaction to the individual company,” he says. Importance Of Communication Communication with employees, tenants, clients, law enforcement, etc. is very important Communication with employees, tenants, clients, law enforcement, etc. is very important and, if done well, all goes smoothly, agrees William Santana Li, Knightscope Chairman and CEO. “Showing up with a 400-pound, 5-foot-tall autonomous robot, deploying it and not telling anyone what is happening is ill advised!” Knightscope also advises potential end users to identify clearly the areas of improvement needed in a security program to guide the deployment of robots. Beware of “Science Fiction Disease,” whose symptoms include unrealistic expectations or fears emanating from Hollywood depictions of robots over the years. Expectations should be spelled out: Keep ongoing and clear communications between the provider and the client, continuing to make improvements together. Future Of Robotics And AI Users should also think clearly through their source of funding, including the second and future years of an implementation. Communication is key, involving stakeholders from the CSO to facilities, purchasing to human resources, finance to the CEO. The future of robotics in corporate America is more than the development of the technology. Given advances in artificial intelligence (AI), sensors and software, the technology is the easy part. Thinking more broadly about how robotics can excel in the corporate environment – and make companies safer – is the next big obstacle on the path to effectively using the powerful technology.

How Do News Reports Influence Opinions About Security?

When security topics become a part of current events, it is usually in a negative light. Security generally only becomes news when it fails, sometimes in a dramatic, high profile and tragic way. However, security failures can also shed light on lessons learned and opportunities to improve. Working toward better security can translate into the purchase of more goods and equipment supplied by our market. For additional insights into the intersection of security and current events, we asked this week’s Expert Panel Roundtable: Good news or bad news? How do news reports and/or current events influence the general public’s opinion of physical security?