3DiVi unveils public deepfake detector for open video analysis
3DiVi has launched a public online demonstration of its deepfake detection tool, allowing users open access to frame-by-frame video analysis for signs of manipulation.
The move opens up technology usually reserved for enterprise settings, offering verification capabilities without the need for specialist software or hardware.
Deepfake challenges
Generative video systems have made it easier for individuals to create deepfake content. As a result, organisations are seeking reliable ways to determine if video submissions have been altered. The 3DiVi Deepfake Detector demo enables uploading of video files or use of a live camera stream, with immediate analysis delivered as each frame is played back.
Tool features
The online demo showcases the company's deepfake detection API module. It lets anyone conduct unlimited tests without registration. The system highlights frames where possible manipulation is detected, providing a visible trace throughout the video's timeline.
This aligns with enterprise requirements, where frame-level analysis is critical for identity verification and compliance processes.
Users have no limit on the number of samples they can check. No installation is required, and the interface highlights any inconsistencies in real time during video playback.
Enterprise applications
The underlying module in the demo supports use cases that extend to identity verification, content moderation, regulatory compliance, and media intake processes. It is intended to be integrated with existing platforms involved in video-based workflows, from onboarding and customer service to evaluating user-generated content.
"As deepfake technologies continue to evolve, it becomes important to evaluate video streams and video files more carefully," said Dmitry Morozov, Chairman of the Board, 3DiVi.
Morozov continued, "In these workflows, the authenticity of the video can be as important as the content itself. By making the demo available online, we offer a straightforward way for teams to test the technology and determine how it may apply to their processes."
Potential users
Typical users include teams involved in identity verification, such as those handling liveness checks in remote onboarding, corporate security departments reviewing internal video communications, and media platforms moderating user submissions.
Software developers building video-centric applications may also benefit from understanding the technology's interpretation of manipulated content.
The demo is intended to help these groups test real video samples and evaluate how the technology responds in various scenarios. This access is positioned to assist with the assessment of the detector's behaviour before potential deployment in production environments.
Transparency and education
The public nature of the demo is being used to promote transparency in the detection process, allowing scrutiny of how frame-level analysis is conducted.
3DiVi also indicated that the tool is designed as an educational resource for developers and researchers examining video forgery detection techniques. The demonstration provides a means of testing workflows and understanding the integration process with existing systems.
According to the company, open access is intended to lower technical barriers and help those who need familiarity with detection technology to begin their evaluation without significant commitment or cost.
