While education authorities in Taiwan and many other countries are promoting unwalled campuses, campus security is emerging as a major concern, and smart video security systems offered by Taiwan's AI startup Umbo Computer Vision can help address the issue, according to Shawn Guan, CEO of the firm.
Guan said that in the past, video security systems were programmed with predefined rules that rely heavily on geometrics to detect motions or changes, and therefore they may work in performing simple tasks in generic environments but fail everywhere else.
Guan stressed that his firm's smart video security systems are available with AI-based recognition function that can tell what actions are going on by judging continuous motions of a person or object, and the systems will issue warning signals to the back-end control center, in case of any suspicious actions such as crossing the wall.
Besides detecting abnormal intrusion into campus, the systems can also detect the behavior of children to see if they are skipping classes, undergoing physical altercations, falling off or getting injured, Guan continued.
Founded in 2014, Umbo has tapped into security control space with its AI-based image recognition technology, but the technology focuses more on analyzing human behavior to ward off potential danger factors, instead of conducting simple facial recognition and identity check.
Umbo's smart video security systems are well received in Taiwan and foreign markets. In Taiwan, the Agency of Corrections under the Ministry of Justice has adopted the systems to judge whether prison inmates show offensive behavior when getting together, and some elementary schools in northern Taiwan have also utilized the systems to boost campus security. In addition, many of the top 500 US enterprises are among the company's customers, according to Guan.