San Francisco, February 6, 2024 — Paravision, a leader in trusted Vision AI for identity applications, today introduced Paravision Deepfake Detection, a cutting-edge solution designed to combat identity fraud and misinformation through advanced AI analysis. Developed with an ethics-first approach, Paravision Deepfake Detection delivers world-class accuracy in a partner-friendly, cloud-ready package, providing a robust defense against the rising threat of digital face manipulations in various sectors.
The accelerating prevalence of deepfaked imagery in traditional and social media poses significant challenges to the public’s understanding of what is real and what is not – with critical implications for foundational concepts of identity, democracy, national security, and human rights. This series of threats is particularly meaningful in the global election year of 2024, compelling leaders in identity, media, and government to prepare to fight AI-powered fraud and misinformation.
The increasing use of deepfakes and digitally manipulated media is concerning, especially as we begin a pivotal election year. Meanwhile, we are seeing a rapid uptick in deepfake-enabled identity fraud, which raises concerns for partners across sectors. In an era where trust is more vital and elusive than ever, Paravision is proud to offer a potent tool in the fight against the deceptive capabilities of advanced deepfake technologies.
Joey Pritikin, Paravision Chief Product Officer
Paravision Deepfake Detection employs AI-based analysis to assess the likelihood of digital manipulation using leading deepfake generation technologies, and outputs a score to guide automated or human-in-the-loop fraud analysis.
Stemming from two years of focused R&D in collaboration with a trusted Five Eyes Government partner, built with a demographically diverse proprietary dataset of over 1 million properly-consented images, and based on Paravision’s core technology and expertise in face image analysis, Paravision Deepfake Detection is designed from the ground-up to deliver outstanding accuracy in mission critical production environments.
With Deepfake Detection, we’ve had the opportunity to leverage so much of what we’ve been building over the last five-plus years: an outstanding technology base, the ability to create truly diverse, properly-consented datasets, a passion for building exceptional accuracy benchmarks, and an understanding of how to efficiently deploy modern, cloud-ready software to our partners’ production environments.
Charlie Rice, Paravision CTO
Paravision recently announced the availability of its Liveness Detection software, enabling the detection of physical presentation attacks such as masks, displays, or printouts using a single image taken from a standard smartphone camera. Paravision Deepfake Detection is an ideal complement to Paravision Liveness, offering the ability to detect digital face manipulations alongside physical spoofs. Both technologies are available for deployment via Paravision’s standard Docker products, marking a significant stride in securing digital landscapes and providing advanced protection against identity fraud.