Release Source:
Release Time:
On December 20, the 2026 National Postgraduate Entrance Examination officially kicked off. Facing the dual challenges of upgraded requirements for intelligent security checks and intelligent patrol inspection, hundreds of smart exam devices have been deployed at multiple test centers, including Xi'an Jiaotong University (Xingqing Campus, Yanta Campus, Innovation Port Campus), Yulin University and Shaanxi University of Chinese Medicine. The solution once again escorts the postgraduate entrance examination, ensuring the exam is conducted in a fair, efficient and orderly manner.


This image is reposted from the Graduate Admissions Office of Xi'an Jiaotong University
Intelligent Security Checkpoint Detection Service
At present, illegal exam cheating using wireless devices is increasing year by year, seriously undermining the authority, impartiality and solemnity of national education examinations and exerting a negative impact on the fair and unbiased education system. Exam fraud has become an unavoidable social problem. Lawbreakers keep updating cheating methods by using high-tech devices such as smartphones and smartwatches, challenging the bottom line of exam management. Our intelligent security gates and other facilities effectively restrict electronic devices from being brought into exam venues.
At the same time, current education exam management still relies mainly on traditional administrative means, while information technology application remains relatively backward. It is difficult to make up for the deficiencies of the traditional management model, and there is a lack of capabilities for active identification and early prevention of irregular behaviors. Relying on advanced technologies such as intelligent patrol inspection and video analysis, it is imperative to shift from post-event accountability to pre-event prevention and in-process control, curb irregularities at the source, upload exam data in real time, and build a fair, impartial and transparent exam environment.


AI Patrol System Based on Visual Algorithms
Abnormal moves and violations by candidates in the venue are identified and captured in real time by the AI system, followed by disposal by on-site exam staff. Similarly, Xi'an Jiaotong University has launched real-time AI intelligent patrol this year, while retaining the "triple insurance" mechanism of manual invigilation, video invigilation and mobile patrol, realizing human-machine collaboration and accurate identification. The system collects real-time images via classroom cameras. With computer vision analysis, AI behavior recognition algorithms, facial recognition algorithms and big data technology, it can accurately identify typical abnormal behaviors such as whispering, looking around, passing items, answering late or starting writing ahead of time.


3D Facial Recognition Terminal
3D facial recognition terminals are well recognized for high accuracy, delivering efficient solutions for smart exam scenarios. They support multiple access control configurations, live detection and person-ID verification to ensure full security. Specially optimized cameras and large-capacity databases guarantee data security. Xitu Light has accumulated profound technical strengths in 3D facial recognition. Leveraging this edge, the company has independently developed full-stack algorithms, modules and intelligent hardware, and built a variety of system-level solutions. These have been widely applied in multiple scenarios including smart mines, intelligent warehousing, smart campuses, smart communities and bank identity verification systems.
This is a vivid footnote to the mission of technology for good and service foremost.Xitu Light upholds every earnest dream through an invisible network of wisdom; it responds to every striving effort with quiet and precise protection. While candidates write quietly with their pens, the smart exam management system runs steadily alongside. Together they compose a present of fairness and order, and shape a future brimming with outstanding talents.

Dedicated to the research and development of core visual algorithm technologies, product innovation and industry applications, empowering AI+ diversified scenarios, facilitating industrial upgrading
Contact Us