Enterprise video authentication consistency verification method and system based on multi-dimensional data fusion

By combining dynamic visual challenges and acoustic interaction commands with multi-dimensional feature extraction and spatiotemporal graph neural network analysis, the problem of existing video authentication technologies being unable to verify the authenticity of scenes has been solved, achieving panoramic consistency authentication of users and scenes and improving the anti-counterfeiting capabilities of remote identity authentication.

CN122049587BActive Publication Date: 2026-07-24BEIJING ZHIQI SHUMEI TECH CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
BEIJING ZHIQI SHUMEI TECH CO LTD
Filing Date
2026-01-30
Publication Date
2026-07-24

AI Technical Summary

Technical Problem

Existing video authentication technologies rely solely on facial liveness detection, which cannot verify the authenticity of key scene elements such as the user's claimed office environment and company logo, or their spatiotemporal consistency with the user's behavior, thus failing to effectively defend against virtual fraud.

Method used

By sending dynamic visual challenge codes and acoustic interaction commands to the client, multi-dimensional feature sequences are extracted in real time, a multi-element spatiotemporal relationship graph is constructed, a spatiotemporal graph neural network is used for deep analysis, and a global spatiotemporal consistency feature vector is generated by combining multi-level consistency verification, and the authentication result is adaptively decided.

Benefits of technology

It achieves comprehensive verification of user authenticity, physical scene authenticity, and logical consistency of interaction process, effectively defends against attacks such as virtual camera injection and deepfake scenarios, and improves the security and reliability of remote identity authentication.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122049587B_ABST
    Figure CN122049587B_ABST
Patent Text Reader

Abstract

The application relates to an enterprise video authentication consistency verification method and system based on multi-dimensional data fusion, which comprises the following steps: issuing a dynamic visual challenge code and an acoustic interaction instruction, guiding a user to perform multi-modal interaction with an authentication physical object, a challenge code and an environment in a real physical space; after synchronously receiving an audio and video stream, a time sequence feature sequence is extracted in real time, continuous frames are abstracted into a multi-element space-time relationship graph sequence, a time sequence graph neural network is used to deeply analyze the space adjacency, motion cooperation and semantic correlation relationship among elements, and a feature vector representing global consistency is output. On this basis, four-dimensional verification is performed in parallel, and finally an adaptive decision model is used to fuse the scores of the four dimensions to output an authentication result. The application realizes a leap from single face feature verification to panoramic relationship verification, can effectively prevent the occurrence of virtual camera injection, pre-recorded video and deep fake scene, and greatly improves the authenticity of video authentication.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Decision-making method and device based on multi-modal semantic alignment, equipment and medium

    CN120954438A

  • Video evidence chain consistency verification method and system based on space-time metadata

    CN121170670A