FB pixel

Alibaba explains development of facial biometric access control system for people wearing masks

 

alibaba

Alibaba has trained an “eye recognition” algorithm along with a mask detection model to perform biometric identification of people wearing respiratory masks, in a process the company has detailed in a blog post.

When government officials put “Ten Strict” control measures into place in Hangzhou city’s Yuhang district on February 3 to prevent the spread of COVID-19, officials were faced with the challenge of how to monitor and enforce a limitation on people leaving their homes to purchase necessities. With most entry and exit records maintained manually, errors and omissions could render the measures ineffective. An automatic access management record with facial recognition, however, would have to identify individuals wearing masks.

The Alibaba Cloud Intelligent database vector team developed the technology, and provided it for free to certain organizations.

During the feature extraction portion of the registration process, two different sets of features are captured; one for the full face, and one for the area around the eyes and forehead. In testing on the Labeled Faces in the Wild (LFW) dataset, the company’s AnalyticDB eye recognition algorithm scored 99.18 percent accuracy. It also scored 93.4 percent against the cfp_fp dataset, and 94.2 percent matching the agedb dataset. Mask detection is performed prior to facial recognition, so that the correct feature set can be matched.

The system performs vector searches, which is supported by both AnalyticDB for MySQL and AnalyticDB for PostgreSQL. The company says that in real-world applications, AnalyticDB can respond to queries of billion of vector data entries within 100 milliseconds. It is used in security projects in multiple cities. Alibaba provides facial recognition for Shanghai’s smart city system.

The facial recognition and eye recognition, and mask detection models, as well as the community resident management solution have all been made open-source by Alibaba, and the company has provided the systems and the AnalyticDB database engine free of charge to organizations running access management applications in support of pandemic containment.

Article Topics

 |   |   |   |   |   |   | 

Latest Biometrics News

 

Deepfake detection upgrade for Sumsub highlights continuous self-improvement

Sumsub has launched an upgrade to its deepfake detection product with instant online self-learning updates to address rapidly evolving fraud…

 

Metalenz debuts under-display camera for payment-grade face authentication

Unlocking a smartphone with your face used to require a camera placed in a notch or a punch hole in…

 

UK regulators pan patchwork policy for law enforcement facial recognition

The UK’s two Biometrics Commissioners shared cautionary observations about the use of facial recognition in law enforcement over the weekend…

 

UK gov’t seeks covert surveillance tech in benefit fraud crackdown

The UK Department for Work and Pensions (DWP) has published a £2 million (US$2.7 million) tender seeking software and hardware…

 

Biometrics in warfare, surveillance raise new oversight challenges

A new Congressional Research Service (CRS) report warns that biometric technologies are moving from routine identity verification into more consequential…

 

Harvard, Linux Foundation launch open-source wallet for selective data sharing

The internet is seeing a wide-scale push towards identity verification and age assurance, but the question remains: how can users…

Comments

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Biometric Market Analysis and Buyer's Guides

Most Viewed This Week

Featured Company

Biometrics Insight, Opinion

Digital ID In-Depth

Biometrics White Papers

Biometrics Events