The name of the base system (platform): | Artificial intelligence (AI, Artificial intelligence, AI) |
Developers: | Microsoft Research |
Technology: | Cybersecurity - Authentication, cybersecurity - Biometric identification, the Systems of video analytics |
Main articles:
- DeepFake
- Artificial intelligence in video
- Video analytics (terms, scopes of application, technologies)
2020: Submission of Microsoft Video Authenticator
The Microsoft company told about the technology developed for fight against dipfeyka who can be used for distribution of misinformation. The tool under the name Microsoft Video Authenticator uses a special algorithm for determination of authenticity of any image or the video. It became known on September 2, 2020.
The technology was developed by the Microsoft Research command in cooperation with Microsoft Responsible AI and Microsoft Committee on AI, ethics and effects in development and researches (Microsoft AI, Ethics and Effects in Engineering and Research, AETHER).
Microsoft Video Authenticator it is capable to perform the analysis of the photo and video content processed using AI algorithms. Dipfeyki is represented by images or video on which after processing what was not actually is shown. The dipfeyka almost indistinguishable from real videos allow to create modern technologies. Similar counterfeits can be used by malefactors for distribution of misinformation, introduction of people to delusion and manipulations.
A system can reveal mixing border a dipfeyka with thin pixels of fading or gray color which are difficult for noticing with the naked eye. During processing of video a system is also capable to estimate authenticity of each separate frame in real time.
This initiative of fight against misinformation is a part of the Microsoft Defending Democracy Program program which also includes such technologies as ElectionGuard for protection of vote and also AccountGuard for protection of the political[1]