Aurora boasts LFW results for facial recognition algorithm

Aurora Computer Services has just announced that its new face recognition algorithm has achieved a mean classification accuracy of 93.24%, based on the LFW benchmark.

According to the company, the LFW benchmark for facial recognition includes difficult use cases like adverse lighting, bad head angles and variations in image quality for its evaluations. Aurora says its reults place is ahead of its competitors in terms of its mean classification score.

Aurora has also announced the launch of its early beta cloud-based face recognition service, which is free to use.

Reported previously, in 2013, Aurora launched a new facial recognition system, aimed at rapid identification. The new system – called Face on the Fly – uses a powerful camera with a wide-angle lens.

Based on a recently-published research report, the global facial recognition market is estimated to grow from $1.92 billion in 2013 to $6.5 billion in 2018, at an impressive CAGR of 27.7%.

BiometricUpdate.com took an extended look at facial recognition as a biometric modality, including public adoption, large-scale deployments and market-sizing. Read the full feature here.

Article Topics

 |   |   | 

Comments

One Reply to “Aurora boasts LFW results for facial recognition algorithm”

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Most Popular

Biometrics Research Group

Biometrics White Papers

Biometrics Events

Explaining Biometrics