Amazon Rekognition Facial Detection Tested

By IPVM Team, Published Dec 03, 2018, 08:48am EST (Research)

******'* *********** ******* *** ********* ************, that ** ********, ***** ** * significant ******* **** ************ ****** *********.

amazon recognition

** **********,****** ******:

****** *********** *** *** ****** ** percent **** ***** – **** ***** have **** ********** ****** – ** images **** **** **** ** *** most *********** ********** ********* *******... ***** aspects ***** *******pose ********** caused by head movement and/or camera movements, ************ ** ********** ** ********** ******* (such ** ***** ******* ** ****, hair, ** ***** ** ******* ****** in *** **********),illumination ********** (such as low contrast and shadows), bright lighting that leads to washed out faces, low ******* and resolution that leads to noisy *** ****** *****, and distortion from cameras and lenses themselves. [emphasis added]

**** ********* * ******* ** ***+ 3-second ***** *****:

facial detection dataset 2

***** ***** *** **** ****** * number ** ****** ****-***** *********:

  • *** ***** (*** ***)
  • *** ***** (* ***)
  • **** **** (**** *° ** **°)
  • **** **** (**** *° ** **°)

** ******** ******* ************ (**** ** 200 ***, **° **** ****, **° down **** ** * ***, **° side ****, **° **** ****, ***.).

** **** *** ***** ** ***** 100+ ***** ****** ****** ********* **********:

  • ****
  • ***
  • ******** (************ *** ****-*********-******-**** ***** ** Intel ** *** ****** * ***** expanded ******** *** **** ****)
  • ****** ***********

******, ** ******* **** ******* *** key ********* ** ***********.

Key ********

****** *********** ****** ********* ******** *** far ****** **** *** ****** ****** including ****** ** **** ** ********:

facial detection overall accuracy

*** ****** *** ********* ***** '***' but **** ** *** ******** ** used **** ***** ****, ********* *** lightning *** *** ******.

*** *******, **** ** ******* *** scores ** *** ********** ****, **** for ****** ***********, ***** *********** ** performance ******* *** *******:

rekognition accuracy segmented

**** ******* *** *** ****** ****** lighting ***, **** '****' ***** *********** at ****** **% ***** *** ***** performance ** ***** **%.

***** *********** *** ********* ** ****** detection ********, ** ***, *** *** away, *** *******, ** *** ****** of ********* ****** ******** *** **** running ** ** **:

facial detection fps

*********** ** **** ****** **** *** chart ******** ** *** *********** ***** service ******** *** ***** ** ** sent ***** ** *** ***** *** other ******** *** ** **** ******* / ** *** **** / ***** the ****** **. ** ******* ***** measurement ******* ** *** ***** ******* of **** ******.

********

*****, ** ******* ** ***% ******** for ****** ***********, ***** ** ****** strong *********** *** **** ** **° down *** **° **** **** **** the ******:

[video-to-gif output image]

****, ** ******* ** **% ******** (meaning ~* *** ** ** ****** correctly ******* * ****) *** ****** Rekognition, ****** * ***** ********* ** steep **** *** **** **** **** the *******'* **** ****:

*** ***** ******, ** *****, **** much **** *********** **** *** ***********. For *******, **** * *** ** every * ****** *** * **** accurately ******** ****** **** * ***** angle *** *** ****** / ********:

*******, *** ** *** *** **** performances *** *********** *** ** **** test ********, ****** *** *****, *** subject ****** ******** ******* *** ****** but ***** **** *** *** ******** at ***:

Accuracy ***********

*** ******** ****** **** ** ************** and * **************. *********** **** ********* evaluation **** * ****** ******"**** ******* *********" (***), ***** ******** ***********-**-******* ** ****** face ********* ******** ************** ** ****** **** ********-*** ******* (something ****** "************ **** *****" ** IoU) *** ******* **** *** ********* captured * ****. ** ****'* ***** library, **** **** ****** *** *** face (***** ***** *** * ****, and ***** ** **** ***). ***** that, ** *** *** * ****** heuristic *** ********: ** *** **** frames *** * **** ******** (******** / *****-***** = ********). **** ** not *******: ** **********, ** **** not ******* *** ***** ********* (* face ** ******** ***********, **** ** the **** ** ** *** *******'* shirt). **** ******* ****** - ** we ***** **** *** *********. ** allows ** **** ******* **** ********** (lux, *****, ***) ******* ****** ** label ******** ***** *** ***** ***** of *** *******.

FPS ***********

*** ******* **** ****** **** *** Rekognition ***:

  1. ****** *** ***** ** ****** **
  2. ***** *** *****
  3. ***** *********** (******** ** ** *** S3 ****, *********** *****->***********)
  4. **** *** *****, ********* ***

* **** ****** ******** *** ************ would ** ** ****** ****** ******** via*******->***********. **** ***** *** *** ********** overhead ** * *****-**-***** *****, ***** our ***** ***** *** ********** ******** from *** *** *********** ********. ** such, *** *** ***********, ******* **** low, ***** ** **** *****.

***** ** **** ************, **** ** a ***** ***** ******* ***********, ** others ** *** ** * *******. For *******, *** *** *** ***-***** hardware *** ****** ******* ***********. *** other ****** ** **** ******* *** on ****** * ** ***** ** devices (****-***) *******, ** **** *********** trade-off ** *********-*********. ************, *********** ** a ******* *******, ***** *********** * model (**** ********'* ****-******-****) ***** **** configuration *** *****. ** ***** ****-**** performance **** *********** ** ******** (****: we *** ***** ************* ** **** is ********), *** ***-****-**** ************, *********** could ** ***** ******.

*******

*** * ****-**** ***** ************ ***********,*********** ************ ** ***** **** *** *** other ************ ***** **** ********** ***** processing ** ******, ** ***** ** far **** ********** **** **-**** **********. For *******, *********** ******* *** **-**** (N. ********) ** $*.** *** * min ** **** ****** ***** ********. Even **** **** **% ******** ** analyze (*.*., *,*** ******* ** * month), *** ******* **** ***** ** ~$50 (*,*** * **% * $*.**) just *** *** **** *********.

** *** ***** ****,****** *********** ****** ******* ***************** ****** ***********, ****** ******** (***, gender, *******), ****** *** ********** *********, etc. **** **** ********* ******** ****** do *** **** ** ***** ** likely ** ******** ** ******* **** their ***********.

********

**** ** *** *** ** ****'* new ****** ** **** ******** *****, starting ********** ********* ******,***** ****** ******* ***** * / Movidius ** ******* *** ****. ** *** ******** a ******* ** *** ***** ** the ******** ******. *** ********, ******** or *********, ****** *** ** *** comments.

Comments (8)

For those of you interested in the fundamentals, check out our new class next week:

deep learning webinar2

Tyler will be explaining fundamentals as part of our parallel effort to provide comprehensive training and education, in addition to our ongoing tests.

Agree
Disagree
Informative: 4
Unhelpful
Funny

Was any testing done running the algorithms with IR nightvision on? Most cameras would switch to IR nightvision (if they have the option) at 1 lux

Agree: 1
Disagree
Informative: 1
Unhelpful
Funny

We did not do any IR versions but we will try some in future tests.

Agree
Disagree
Informative
Unhelpful
Funny

Just wait until they integrate it with Ring if they haven't already.

Agree: 1
Disagree
Informative
Unhelpful
Funny

Is there any way you can post all of your raw video used in the tests for IPVM users to download and run thru their own Analytics systems?

Agree: 2
Disagree
Informative
Unhelpful
Funny

I think this would be a really great way to compare products.

Agree: 1
Disagree
Informative
Unhelpful
Funny

We might release the dataset but have not decided yet.

Agree
Disagree
Informative: 1
Unhelpful
Funny

Even with just 10% activity to analyze (i.e., 4,380 minutes in a month), the monthly cost would be ~$50 (4,380 x 10% x $0.12) just for the face analytics.

Aren't you considering the "10%" factor twice in this calculation? 4,380 minutes itself is 10% of total minutes in a month.

Agree: 1
Disagree
Informative
Unhelpful
Funny
Subscribe to IPVM Research to read the full report.
Why do I need to subscribe?
The IPVM Research Service includes products tests and shootouts plus competitive and financial analysis, helping decision-makers better evaluate purchasing, partnering, developing, and/or competing against companies in physical security.
Already have an account?
Loading Related Reports