Member Discussion

Trueface, Facefirst, Anyvision Facial Recognition Bias

Robert J. Miller

06/18/20 01:46pm

What do companies like TrueFace, Anyvision and FaceFirst say about the bias towards people of color and their algorithms? They have been very quiet.

Login to read this IPVM discussion.
Why do I need to log in?
IPVM conducts unique testing and research funded by member's payments enabling us to offer the most independent, accurate and in-depth information.

Undisclosed #1

06/19/20 04:10am

****** **** *** *** asking *** * ***% proof ** *******.

*** ** *** ******** world **** **** ** to **** *********?

* **** * ****.

Robert J. Miller

06/19/20 01:40pm

****** **** ****** *********** bias ** ******* **** and **** ** ******** facial ***********, ******* *** these ***** ******* ***** stuck ** *** ***** daily, *** ******** ** bias **** ***** ****** recognition ********* **** ***** us ** *********** ** answer ********* ********* **** clients.

** **** * **** / ** ******* ** these *************** ********* ****.

Avatar

Brent Boekestein

06/19/20 07:58pm

*** *******, ******, ********* some **** ***** *** methodology *** ******* **** to ****** **** ** ML ******* **** ****. You *** **** **** about ** **** -*&* | ******, ***.. ** **** **** work ** ** *** progress *** ** **** in ******** **** ** companies *** ******* ** make ** * ********. I'd **** ** **** from *** *********. *** customers ********** ** **** rec ****** ********** ********* about ********* *********** ****** skin ***** *** ******?

Avatar

Shaun Moore

09/20/20 03:27pm

@****** - **, **********,**** ********* * *** pieces ** ******* ** bias **** *** ** viewed *****.

**** ********** ** **- ****** ******** ****** Meetup

******** **** **** ********

****** - ******* **** Mitigation- * ******* **** of ********* ** **** mitigation ******** *** **/** researchers *** *********.

Undisclosed Manufacturer #2

09/21/20 12:58pm

*** ** ** ****** because ** ****** ** probability. ** **** ***, Face *** ************** (****) allow *** *** * confidence ****** *** *** customer ** ******. ** with * **** ********** and *** **** *** many *******, ** **** too *** ** * confidence ******* *** *** will *** *** **** images.

John Honovich

IPVM | In reply to Undisclosed Manufacturer #2 | 09/21/20 01:05pm

*** ** ** ****** because ** ****** ** probability

********. *** *** ****** is *** ****** *** how **** ** *****, which *** **** ************ amongst *******.

Undisclosed Manufacturer #3

In reply to John Honovich | 09/21/20 02:28pm

******. ** ** ****** because ** **** ****** data. **** ** *** classic "******* ** - garbage ***" *******.

**** ***** **** ****** haven ***** **** ************ ****. **** ***** to ***** *** * great *******. ***** **** a "******* ****" ** the ***/*** ** * color-calibration ********* *** ***** cameras ***** ******** * middle-aged ***** *****. ** a ******, *****'* ***** performance ******* ******* ******* subjects *** ****** **** performance ********. **'** ****** a ******* ******* **** datasets ***** ******* ** train ****** ****. **** if **** *** * diverse *** ** ***** to ***** ***** ****** nets, **** ***** **** to ****** * *******. This ******** ********** *********** the ********.

Top Comment Votes - Past 3 Months
  • 1. John Honovich - 1452
  • 2. Brian Karas - 364
  • 3. Michael Miller - 341
  • 4. Greg Cortina - 303
  • 5. Nick Giannakis - 264
  • 6. Chris Dearing - 224
  • 7. Andrew Myers - 217
  • 8. Brian Rhodes - 196
  • 9. Marty Major - 191
  • 10. Hans Kahler - 167
  • 11. Ethan Ace - 138
  • 12. Shannon Davis - 134
  • 13. Ross Vander Klok - 120
  • 14. Austin Rich - 118
  • 15. Michael Silva - 108
  • 16. Dwayne Cooney - 108
  • 17. Hash Salehi - 90
  • 18. Sean Nelson - 86
  • 19. Anthony Jones - 78
  • 20. Daniel Lewkovitz - 77
Manufacturers
Tools

BestMatch

Camera Calculator

Online Shows

Camera Finder

F-Stop Calculator

Chat

Integrator Finder

News Spider

Camera Comparison

Quizzes

Pages

About

Contact