NVIDIA Unveils Technology to Detect Fake Videos at Lightning Speed

Global tech company NVIDIA has announced the launch of a new advanced tool powered by artificial intelligence that can detect and identify fake videos at an ultra-fast speed of just 22 milliseconds, in a move aimed at curbing the rapid spread of AI-generated deceptive visual content.

The company explained that this innovative technology is not intended to replace traditional journalistic verification processes or digital forensic analyses, but rather designed to serve as an additional support for media organizations and companies, enabling them to instantly verify the authenticity and accuracy of videos before adopting and publishing them to the public.

The new tool is available as part of the NVIDIA NIM Microservices package, giving organizations the advantage of seamless and flexible integration within their existing infrastructure systems and workflows, without incurring the costs of building new dedicated platforms for content review.

The tool's mechanism relies on a frame-by-frame video analysis strategy, then assigns each clip a numerical probability score indicating the likelihood that it was generated or altered using AI technologies. The tool can analyze 1080p videos in just 22 milliseconds when running on RTX graphics cards, making it the ideal solution for real-time, live verification within media production environments.

The company confirmed that its new release topped the results of the 'AI GVD Bench,' the global benchmark specialized in measuring and evaluating the performance of fake video detection systems. The technology outperformed a wide range of competing software solutions, both open-source and commercial, in rigorous tests involving visual materials generated by various and diverse AI models.