Welcome to DU!
The truly grassroots left-of-center political community where regular people, not algorithms, drive the discussions and set the standards.
Join the community:
Create a free account
Support DU (and get rid of ads!):
Become a Star Member
Latest Breaking News
General Discussion
The DU Lounge
All Forums
Issue Forums
Culture Forums
Alliance Forums
Region Forums
Support Forums
Help & Search
General Discussion
Related: Editorials & Other Articles, Issue Forums, Alliance Forums, Region ForumsAmazon's facial recognition matched 28 members of Congress to criminal mugshots
Amazons facial recognition matched 28 members of Congress to criminal mugshots
New ACLU test illustrates the limits of Amazons Rekognition system
Illustration by James Bareham / The Verge
By Russell Brandom@russellbrandom Jul 26, 2018, 8:02am EDT
The American Civil Liberties Union tested Amazons facial recognition system and the results were not good. To test the systems accuracy, the ACLU scanned the faces of all 535 members of congress against 25,000 public mugshots, using Amazons open Rekognition API. None of the members of Congress were in the mugshot lineup, but Amazons system generated 28 false matches, a finding that the ACLU says raises serious concerns about Rekognitions use by police.
An identification whether accurate or not could cost people their freedom or even their lives, the group said in an accompanying statement. Congress must take these threats seriously, hit the brakes, and enact a moratorium on law enforcement use of face recognition.
Reached by The Verge, an Amazon spokesperson attributed the results to poor calibration. The ACLUs tests were performed using Rekognitions default confidence threshold of 80 percent but Amazon says it recommends at least a 95 percent threshold for law enforcement applications where a false ID might have more significant consequences.
While 80% confidence is an acceptable threshold for photos of hot dogs, chairs, animals, or other social media use cases, the representative said, it wouldnt be appropriate for identifying individuals with a reasonable level of certainty. Still, Rekognition does not enforce that recommendation during the setup process, and theres nothing to prevent law enforcement agencies from using the default setting.
...
New ACLU test illustrates the limits of Amazons Rekognition system
Illustration by James Bareham / The Verge
By Russell Brandom@russellbrandom Jul 26, 2018, 8:02am EDT
The American Civil Liberties Union tested Amazons facial recognition system and the results were not good. To test the systems accuracy, the ACLU scanned the faces of all 535 members of congress against 25,000 public mugshots, using Amazons open Rekognition API. None of the members of Congress were in the mugshot lineup, but Amazons system generated 28 false matches, a finding that the ACLU says raises serious concerns about Rekognitions use by police.
An identification whether accurate or not could cost people their freedom or even their lives, the group said in an accompanying statement. Congress must take these threats seriously, hit the brakes, and enact a moratorium on law enforcement use of face recognition.
Reached by The Verge, an Amazon spokesperson attributed the results to poor calibration. The ACLUs tests were performed using Rekognitions default confidence threshold of 80 percent but Amazon says it recommends at least a 95 percent threshold for law enforcement applications where a false ID might have more significant consequences.
While 80% confidence is an acceptable threshold for photos of hot dogs, chairs, animals, or other social media use cases, the representative said, it wouldnt be appropriate for identifying individuals with a reasonable level of certainty. Still, Rekognition does not enforce that recommendation during the setup process, and theres nothing to prevent law enforcement agencies from using the default setting.
...
More at link
InfoView thread info, including edit history
TrashPut this thread in your Trash Can (My DU » Trash Can)
BookmarkAdd this thread to your Bookmarks (My DU » Bookmarks)
5 replies, 789 views
ShareGet links to this post and/or share on social media
AlertAlert this post for a rule violation
PowersThere are no powers you can use on this post
EditCannot edit other people's posts
ReplyReply to this post
EditCannot edit other people's posts
Rec (7)
ReplyReply to this post
5 replies
= new reply since forum marked as read
Highlight:
NoneDon't highlight anything
5 newestHighlight 5 most recent replies
Amazon's facial recognition matched 28 members of Congress to criminal mugshots (Original Post)
sl8
Jul 2018
OP
lagomorph777
(30,613 posts)1. So Kobach shouldn't use it for voter purges? Or...
The hits were accurate...
fleur-de-lisa
(14,628 posts)2. Only 28? Should have been 248.
blugbox
(951 posts)3. Wow, could you imagine being convicted
because a police computer gave a B- guess that it was you?
If Amazon created the product, why would the police force use a lower threshold than they recommend??
eppur_se_muova
(36,289 posts)5. quelle surprise nt