By Clearview AI BLOG
In the ever-changing realm of technology, artificial intelligence (AI) and its applications have fundamentally transformed numerous sectors, with facial recognition emerging as a critical development. Facial recognition technology has evolved rapidly in recent years, increasing dramatically in accuracy. This blog post will delve into a quick overview and comparison of Clearview AI, and Vigilant FaceSearch (which is owned by Motorola).
THE ALGORITHM
Clearview AI operates on an advanced algorithm capable of identifying faces from multiple angles, eliminating the need to rotate or skew images. The algorithm’s accuracy facilitates a more user-friendly experience and yields more accurate results.
Vigilant FaceSearch, a product of Motorola, has a less capable algorithm. It requires image rotation and skewing to search for off-angle photos. While this might add an extra layer of control over image manipulation, it also introduces an additional step in the process.
Clearview AI has consistently ranked high in NIST evaluations across all categories, boasting a remarkable #2 position globally and #1 in the U.S. for WILD photos, which signifies its adeptness at identifying faces in uncontrolled environments. The 1:N mugshot category particularly stands out, with an accuracy of 99.85% for picking a photo out of a lineup of 12 million photos.
A notable point of difference lies in the NIST rankings, where Vigilant's algorithm does not feature in the top 100 algorithms for any of the categories. The last data update was in 2019, and there is no indication of active product development or investment.
THE DATABASE
When using Vigilant FaceSearch, your gallery size is limited to how many images you may have access to in your local agency, and a small pre-populated dataset of arrest records provided by Vigilant FaceSearch. By contrast, Clearview AI's extensive public image database has over 30 billion images and continues to grow. With Clearview AI, you are also able to also search your own local images, so there is nothing that Vigilant FaceSearch can do that Clearview AI can not do.
Accuracy and gallery size are very important to a facial recognition program. A larger dataset is more likely to help you find lead information in investigating crimes , and a larger dataset also helps prevent possible wrongful identifications.
AUDITING & COMPLIANCE
Clearview AI is specifically designed for government and public safety use cases and is committed to complying with government guidelines on facial recognition. Clearview AI has one of the most comprehensive permission systems, auditing capabilities, workflows, admin tools and compliance features in the industry.
Clearview AI’s technology is proven and has been SOC 2 certified for its cybersecurity and data handling practices. Even though Vigilant FaceSearch is a cloud based product, it has not been audited by an independent third party for security.
THE COMPANIES
Vigilant FaceSearch is part of Motorola Solutions, a globally recognized brand with a long history in telecommunications and electronics. Despite the parent company's size and reputation, Vigilant FaceSearch's development seems to have taken a back seat. The last significant update was in 2019, and there has not been any notable product investment or enhancement since then. This stagnation stands in stark contrast to the rapid advancements we see in today's tech landscape. The user interface is reminiscent of older systems, and compared to modern counterparts, can be perceived as less intuitive and more complex.
On the other hand, Clearview AI is an independent company, fully focused on our core facial recognition product. We are always striving to improve our technology, increase our dataset, and improve our accuracy. Our company is committed to provide cutting edge technology for our government and law enforcement customers, to help them with their mission in solving crime.