99.9997% Accurate Amazon Facial Rekognition Falsely Matches 28

By: John Honovich, Published on Jul 27, 2018

A new facial recognition controversy has arisen, with the ACLU cleverly using Amazon's Rekognition service to falsely match 28 Congresspeople to mugshot photos.

However, the ACLU's results also show that the Amazon system was 99.9997% accurate and yet still generated those 28 false matches. While that may seem paradoxical, let us examine how that works.

99.9997% Accuracy

The ACLU test required Amazon Rekognition to conduct more than 13 million facial comparisons as it submitted the 535 members of Congress vs a 25,000 mugshot database. Amazon needs to compare all the combinations to determine whether any of them are a 'match', as our graphic below shows:

Out of those 13+ million comparisons, 28 of them were wrong, with Congresspeople being falsely identified as having been arrested [insert joke about Congresspeople being criminals, etc.]. The reality though is that in 99.9997% of the comparisons, Amazon made the correct decision, as the graphic below illustrates:

This is not a defense of Amazon; Rather it is to emphasize the fundamental challenges of doing facial recognition at a large scale. AI expert Andrew Ng explains this in the video embedded below:

Could Be Worse Actually

Get Video Surveillance News In Your Inbox
Get Video Surveillance News In Your Inbox

Indeed, the results of the ACLU test would have been worse if (1) the arrested people were of similar ages to the Congresspeople (i.e., the average Congressperson is ~60, the average arrested person's age is significantly less) and (2) surveillance video images were used (i.e., all images in this test were cooperative portrait style photos taken at direct angles, much easier to compare).

Amazon's Defense Ignores False Negatives

Amazon was quick to defend their facial recognition software, telling the Verge:

[Amazon] recommends at least a 95 percent threshold for law enforcement applications where a false ID might have more significant consequences. [emphasis added]

However, a major problem of raising it from the default 80% that the ACLU test used is that going to 95% radically increases the number of false negatives, that is the true matches that are not returned because the system is set at such a high threshold.

It is easy to demonstrate this with Amazon's Rekognition own demo, as the following examples we ran show how they would incorrectly not match at 95%:

At 95%, all of these valid true matches would be rejected by the system because their scores would be too low.

Fundamental Challenges Remain

At scale, false positives are going to be a significant risk, both for operationally managing them and for the risks to those people falsely matched. And eliminating those false positives, by raising the similarity threshold, will ensure many true positives are missed, i.e., false negatives, undermining the utility of the system. These challenges will remain and are at the core of increasing public debates about how this technology should be used.

1 report cite this report:

RealNetworks Free School Facial Recognition on Aug 03, 2018
The company that created RealPlayer is moving beyond media delivery and into the security space with a new facial recognition platform they have...
Comments (19) : PRO Members only. Login. or Join.

Most Recent Industry Reports

Verkada Video Quality Problems Tested on May 23, 2019
Verkada suffers from numerous video quality problems, not found in commercial IP cameras, new IPVM testing of Verkada vs Axis and Hikvision...
Average Frame Rate Video Surveillance 2019 on May 23, 2019
What is the average frame rated used in video surveillance systems? In IPVM's 2011 statistics, the average was 6-8fps increasing to ~10fps in...
Access Control Job Walk Guide on May 22, 2019
Significant money can be saved and problems avoided with an access control job walk if you know what to look for and what to ask. By inviting...
US Considers Sanctions Against Hikvision and Dahua on May 22, 2019
The US government is considering blacklisting "up to 5" PRC surveillance firms, including Hikvision and Dahua, Bloomberg reported, with human...
Dahua USA Celebrates 5 Years of Errors on May 21, 2019
Dahua USA is, in their own words, 'celebrating' 5 years in North America or as trade magazine SSN declared: Dahua Technology finds success in...
Axis ~$150 Outdoor Camera Tested on May 21, 2019
Axis has released the latest in their Companion camera line, the outdoor Companion Dome Mini LE, a 1080p integrated IR model aiming to compete with...
Covert Facial Recognition Using Axis and Amazon By NYTimes on May 20, 2019
What if you took a 33MP Axis camera covering one of the busiest parks in the US and ran Amazon Facial Recognition against it? That is what the...
Amazon Ring Public Subsidy Program Aims To Dominate Residential Security on May 20, 2019
Amazon dominates market after market. Quitely, but increasingly, they are doing so in residential security, through a combination of significant...
LifeSafety Power NetLink Vulnerabilities And Problematic Response on May 20, 2019
'Power supplies' are not devices that many think about when considering vulnerabilities but as more and more devices go 'online', the risks for...

The world's leading video surveillance information source, IPVM provides the best reporting, testing and training for 10,000+ members globally. Dedicated to independent and objective information, we uniquely refuse any and all advertisements, sponsorship and consulting from manufacturers.

About | FAQ | Contact