As a member of the FIDO Alliance, CyberLink commits to developing and supporting new biometrics-based authentication standards through FaceMe®, its facial recognition technology

TAIPEI, TAIWAN — April 21, 2022 — CyberLink Corp. (5203.TW), a pioneer of AI and facial recognition technologies, announces it became an Associate Member of the FIDO Alliance, an open industry association focused on authentication standards and device attestation. CyberLink joins hundreds of industry players in developing and implementing new authentication specifications and standards, including facial recognition, that will better protect user privacy.

The FIDO Alliance is an industry association centered on advancing authentication standards to decrease dependence on passwords, which can be easily stolen or hacked. These new standards make authentication simpler for consumers to use and easier for service providers to manage. The Alliance’s rich membership comprises stakeholders that include government agencies, service providers, technology players and financial services industry leaders.

Technology industry member CyberLink is dedicated to providing encrypted biometric authentication through its FaceMe AI facial recognition solutions. With its facial recognition expertise, CyberLink is well-positioned to help the Alliance promote the adoption of biometric technologies that will deliver to our society more secure and convenient authentication alternatives to those with inherent flaws that are generally used today.

“We are excited to welcome our newest Associate Member CyberLink,” said Andrew Shikiar, Executive Director and CMO of the FIDO Alliance. “The FIDO vision of universal strong authentication promises better security, enhanced privacy, more commerce and expansion of services throughout digital industries. CyberLink’s addition to our Alliance supports our industry goal to make user authentication easier and safer for all parties.”

“As privacy and protection are increasingly critical imperatives of our society’s reliance on technology in every facet of life, users deserve stronger yet more intuitive authentication methods,” said Dr. Jau Huang, CEO of CyberLink. “I have no doubt that CyberLink’s biometrics authentication expertise, cemented through FaceMe, our facial recognition technology, will benefit the FIDO Alliance in establishing and providing safer authentication methods for end-users.”

旗下新品FaceMe® Platform人臉辨識API平台首次亮相
開發人員可快速部署多樣化的人臉辨識解決方案，提升導入效率

【2022年3月16日，台北訊】全球頂尖臉部辨識開發商訊連科技將於「2022智慧城市展 Smart City Summit & Expo」展出智慧安控及門禁、智慧物聯網及智慧金融等多項採用旗下FaceMe® AI人臉辨識的各種智慧解決方案。「2022智慧城市展」將於3月22日至3月25日於南港展覽館2館展出，訊連科技攤位為P104。

隨著邊緣AI技術普及，AI人臉辨識亦日趨廣泛，而其中門禁安控更是人臉辨識的主要應用場域。訊連科技推出之FaceMe® Security就是專為安控及門禁設計的人臉辨識解決方案，提供刷臉打卡、口罩與體溫監測、單張相片以圖搜人、影像即時監控回放等功能。FaceMe® Security可相容於現行IP安控架構，如IP攝影機、工作站、NVR等。此外，亦可與Milestone、Network Optix、AXIS、Genetech、AVIGILION及VIVOTEK等主流影像管理系統(VMS)整合，提供即時示警功能。

此次訊連科技展出亮點，還包含首次對外亮相的FaceMe® Platform人臉辨識API平台。FaceMe® Platform主打彈性部署、高可用性及效能，提供用戶各項開發人臉辨識應用所需之核心功能，開發商可利用平台上的開放API介接其服務。訊連科技亦將同時展出專為智慧辦公室設計之FaceMe® TimeClock人臉辨識考勤系統，及專為零售店開發，可提供人流分析之FaceMe® Smart Retail等兩款採用FaceMe® Platform之全新解決方案。

此外，訊連科技亦提供FaceMe® SDK人臉辨識引擎，可支援Windows、Linux、Android、JetPack等平台，讓門禁機開發商於各式硬體中快速整合人臉辨識功能。FaceMe® SDK符合iBeta 依ISO/IEC 30107-3標準進行的活體冒用攻擊Level 2測試，可有效杜絕臉部照片、影片及3D列印面具之活體冒用攻擊。

欲了解AI人臉辨識在智慧安防領域之應用，歡迎於4/27至4/29的Secutech博覽會造訪訊連科技攤位。

Secutech 2022台北國際安全科技應用博覽會

• 活動期間：111年 4 月 27 日 至 4 月 29 日
• 展出時間：上午9:30至下午17:30 （最後一日參觀至下午16:30）
• 活動地點：台北南港展覽館2館1樓（115台北市南港區經貿二路2號）
• 訊連科技攤位編號：826

The end-to-end integration brings FaceMe’s industry-leading facial recognition functionalities to AXIS camera and VMS setups, fusing live stream monitoring and group tagging in a seamless solution

TAIPEI, TAIWAN — April 15, 2022 — CyberLink Corp. (5203.TW), a pioneer of AI and facial recognition technologies, today announced an important update to its facial recognition security software, FaceMe® Security. FaceMe Security now integrates with AXIS Communications’ AXIS Camera Station VMS (video management software), creating a robust, unified platform that brings facial recognition to AXIS’s security and access control applications. With this solution, users can now layer person-of-interest detection, group tagging and management, and visitor summaries on top of existing IP surveillance configurations with AXIS cameras and VMS.

AXIS is a global market leader in intelligent security solutions, integrating network cameras, video recorders, workstations, and video management systems. AXIS Camera Station is an easy-to-use, efficient VMS platform offering video surveillance and access control applications. Its intuitive interface and extensive compatibility make it an excellent fit for FaceMe® Security, a value-added smart surveillance software solution powered by CyberLink’s highly-ranked and precise FaceMe® AI facial recognition engine.

The unified platform, combining FaceMe’s real-time facial recognition to the AXIS Camera Station interface, delivers a unique, efficient and powerful monitoring solution. One operator can visualize up to 36 camera feeds on one screen. They can then rely on the facial recognition console running on a second display to receive real-time alerts. These alerts are triggered when individuals associated to a managed group (e.g., VIP, blocklists, employees) appear in front of a connected camera located in an area or at a time potentially requiring a certain action to be taken.

“Smart video management software is becoming an essential component to security and access control systems,” said Dr. Jau Huang, CEO of CyberLink. “Adding the facial recognition capabilities of FaceMe to the AXIS Camera Station enables a powerful, highly automated security monitoring environment, running on one intuitive platform.”

For an introduction to the FaceMe®-AXIS integration’s new features, visit the FaceMe® Security website or watch the introduction video.

【2022年3月22日，台北訊】AI 人臉辨識領導廠商訊連科技（5203.TW）宣布旗下 FaceMe® 人臉辨識引擎之活體防偽技術，通過 iBeta 的 PAD 測試 Level 2（Presentation Attack Detection，活體冒用攻擊）。iBeta Level 2 測試係透過高精度之 3D 面具進行人臉辨識技術之活體冒用攻擊。訊連科技 FaceMe® 於該項測試中，以 0% 之攻擊破解率（意即可 100% 之防止活體冒用）之佳績通過測試，證明 FaceMe® 為適用於金融、網路交易、遠距身分認證等安全係數要求較高之生物辨識解決方案。

iBeta 為全球少數獲得美國國家標準暨技術研究院 (NIST, U.S. National Institute of Standards and Technology) 之國家實驗室自願認可程序 (NVLAP, National Voluntary Laboratory Accreditation Program) 認證的生物特徵測試實驗室。iBeta 的 PAD 測試是依據 ISO/IEC 30107-3 標準進行的活體冒用攻擊測試，為人臉辨識防偽技術的全球產業公認標準，測試結果深受市場信賴。

FaceMe® 於去 (2021) 年下半年通過 iBeta PAD Level 1 測試，該測試主要以 2D 照片與影片進行。而此次 Level 2 測試則以更精細的 3D 列印面具、樹脂面具、乳膠面具等，進行活體冒用攻擊測試。FaceMe® 的攻擊遭破解率為 0% (APCER, Attack Presentation Classification Error Rate)，代表 FaceMe® 能有效阻擋任何形式之 3D 偽造。此外，FaceMe® 的真人拒認率 (BPCER, Bona Fide Presentation Classification Error Rate) 在 iOS 裝置上僅 1.5%、在 Android 裝置上僅 2.5%。換言之，當測試標的為真人時，FaceMe® 可分別達到 98.5% 與  97.5% 的成功辨識率，以生物辨識詐欺防範而言，為市面上最優秀的人臉辨識解決方案。穩健的活體辨識與防偽技術，在需要遠端認證用戶身分，如：使用網路銀行、購買年齡限制商品，或居家辦公連到公司內網時，尤為重要。

「生物辨識詐欺防範，是金融科技等領域的當務之急；任何安全漏洞都可能造成災難性的影響。此外，用戶要從遠端存取機敏資料，也讓遠端身分驗證的需求爆炸性增長。」訊連科技董事長兼執行長黃肇雄表示。「iBeta PAD 是人臉辨識領域中，廣受全球肯定的活體防偽測試，一向被視為國際標準。FaceMe® 這次以完美的成績通過 Level 2 測試，意味著我們是頂尖的人臉辨識解決方案，能有效防止生物辨識詐欺情形，即便碰上最複雜的 3D 面具攻擊也不成問題。」

FaceMe® 可於各式硬體裝置上運行，包括高階工作站，或 IoT、AIoT 裝置中常見的低功耗晶片裝置皆可導入，是市面上功能最全面且易用的人臉辨識解決方案。FaceMe® 讓系統整合商得以更快速、可靠、精準且彈性的方式，將人臉辨識技術導入門禁安控、公共安全、金融科技、智慧零售或智慧居家等不同應用場景中。

CyberLink FaceMe® iBeta 活體冒用攻擊 (Presentation Attack Detection) 測試報告請參閱：

https://www.ibeta.com/wp-content/uploads/2022/03/220304-CyberLink-PAD-Level-2-Confirmation-Letter.pdf

TAIPEI, TAIWAN — March 16, 2022 — CyberLink Corp. (5203.TW), a pioneer in multimedia and AI technologies, announced today Dr. Jau Haung, its founder and CEO, has been invited to speak at NVIDIA’s GPU Technology Conference 2022 (GTC). Dr. Huang will address the role of AI in helping content creators enhance productivity and improve quality when making engaging, professional videos. The presentation will take place virtually on Thursday, March 24 at 11 a.m. CST (Wednesday, March 23 at 11 p.m. ET/8 p.m. PT).

GTC is a global AI conference held every half year, bringing together hundreds of developers who share the latest trends and applications for AI, computer graphics, data science, and machine learning.

At GTC 2022, Dr. Huang will share how CyberLink utilizes AI-powered tools in PowerDirector, its flagship video editing software, for features such as AI style transfer, sky replacement, motion tracking, and a series of enhancer tools. CyberLink first introduced its video editing software in 2001 and closely followed rising trends for more than two decades – from SD and HD to VR and 4K, from optical discs and files to online and streaming media – before deploying AI for editing in 2020.

“Artificial intelligence changes every aspect of life, from industrial-level image recognition to video analytics, security, autonomous vehicles and digital content creation,” said Dr. Jau Huang, CEO of CyberLink. “As the world-leading AI and multimedia software developer, we use AI extensively in our FaceMe® facial recognition solution as well as our creative software and apps, such as PowerDirector. It is CyberLink’s mission to help video creators at all levels achieve image and audio superiority, allowing them to create something extraordinary no matter their skill level.”

In the past, the use of motion tracking or motion masks was a labor-intensive and time-consuming undertaking reserved to professional creators. With the help of AI technology, PowerDirector delivers fast, accurate motion tracking – with just a few clicks, making this powerful tool accessible to every creator and for all creations, ultimately raising standards across all video productions.

In the same vein, AI brings some of the most sophisticated special effects to novice creators while cutting down dramatically the time spent by industry professionals. For example, PowerDirector’s sky replacement feature uses AI to precisely segment the sky in a video sequence and replace it with footage of sky conditions that better match the desired effect or visual aesthetics. This feature has the power to transform the atmosphere of a scene in a few clicks. From a practical aspect, it enables outdoor filming in a much wider range of sky conditions, eliminating costly interruption from changing weather or simply waiting for the right time between sunrise and sunset.

To learn more about PowerDirector, visit CyberLink online at http://www.cyberlink.com, or sign-up for CyberLink’s track at GTC 2022.

CyberLink @ GTC 2022 Information

• Topic: AI-powered Video Editing for Creators at All Levels [S41616]
• Speaker: Dr. Jau Huang, Founder and CEO of CyberLink
• Time:
  • Thursday, March 24 11:00AM – 11:25AM (CST, Taiwan)
  • Wednesday, March 23 11:00PM – 11:25PM (EDT, New York)
  • Wednesday, March 23 8:00PM – 8:25PM (PDT, Los Angeles)
• GTC 2022 Official Website: https://www.nvidia.com/gtc
• Sing up for CyberLink Track