Amazon Rekognition Facial Detection Tested

By: IPVM Team, Published on Dec 03, 2018

Amazon's Rekognition service has announced improvements, that if accurate, would be a significant advance over conventional facial detection.

amazon recognition

In particular, Amazon claims:

Amazon Rekognition can now detect 40 percent more faces – that would have been previously missed – in images that have some of the most challenging conditions described earlier... These aspects might include pose variations caused by head movement and/or camera movements, occlusion due to foreground or background objects (such as faces covered by hats, hair, or hands of another person in the foreground), illumination variations (such as low contrast and shadows), bright lighting that leads to washed out faces, low quality and resolution that leads to noisy and blurry faces, and distortion from cameras and lenses themselves. [emphasis added]

IPVM developed a dataset of 100+ 3-second video clips:

facial detection dataset 2

These clips are done across a number of common real-world scenarios:

  • Day light (200 lux)
  • Low light (1 lux)
  • Side tilt (from 0° to 90°)
  • Down tilt (from 0° to 30°)

We compiled various combinations (such as 200 lux, 15° side tilt, 22° down tilt or 1 lux, 75° side tilt, 30° down tilt, etc.).

We then ran tests of these 100+ clips across facial detection approaches:

  • HAAR
  • HOG
  • Openvino (specifically the face-detection-retail-0004 model on Intel i7 and Myriad X chips expanded upon from our NCS2 test)
  • Amazon Rekognition

Inside, we explain full results and key variances in performance.

******'* *********** ******* *** announced ************, **** ** accurate, ***** ** * significant ******* **** ************ facial *********.

amazon recognition

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

****** *********** *** *** detect ** ******* **** faces – **** ***** have **** ********** ****** ** ****** **** have **** ** *** most *********** ********** ********* earlier... ***** ******* ***** includepose ********** caused by head movement and/or camera movements, ************ ** ********** ** background ******* (**** ** faces ******* ** ****, hair, ** ***** ** another ****** ** *** foreground),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]

**** ********* * ******* of ***+ *-****** ***** clips:

facial detection dataset 2

***** ***** *** **** across * ****** ** common ****-***** *********:

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

** ******** ******* ************ (such ** *** ***, 15° **** ****, **° down **** ** * lux, **° **** ****, 30° **** ****, ***.).

** **** *** ***** of ***** ***+ ***** across ****** ********* **********:

  • ****
  • ***
  • ******** (************ *** ****-*********-******-**** model ** ***** ** and ****** * ***** expanded ******** *** **** ****)
  • ****** ***********

******, ** ******* **** results *** *** ********* in ***********.

[***************]

Key ********

****** *********** ****** ********* accuracy *** *** ****** than *** ****** ****** including ****** ** **** of ********:

facial detection overall accuracy

*** ****** *** ********* quite '***' *** **** of *** ******** ** used **** ***** ****, including *** ********* *** bad ******.

*** *******, **** ** segment *** ****** ** key ********** ****, **** for ****** ***********, ***** differences ** *********** ******* are *******:

rekognition accuracy segmented

**** ******* *** *** severe ****** ******** ***, with '****' ***** *********** at ****** **% ***** low ***** *********** ** under **%.

***** *********** *** ********* in ****** ********* ********, it ***, *** *** away, *** *******, ** two ****** ** ********* versus ******** *** **** running ** ** **:

facial detection fps

*********** ** **** ****** than *** ***** ******** as *** *********** ***** service ******** *** ***** to ** **** ***** to *** ***** *** other ******** *** ** done ******* / ** the **** / ***** the ****** **. ** examine ***** *********** ******* in *** ***** ******* of **** ******.

********

*****, ** ******* ** 100% ******** *** ****** Rekognition, ***** ** ****** strong *********** *** **** is **° **** *** 45° **** **** **** the ******:

[video-to-gif output image]

****, ** ******* ** 59% ******** (******* ~* out ** ** ****** correctly ******* * ****) for ****** ***********, ****** a ***** ********* ** steep **** *** **** tilt **** *** *******'* head ****:

*** ***** ******, ** noted, **** **** **** challenging **** *** ***********. For *******, **** * out ** ***** * frames *** * **** accurately ******** ****** **** a ***** ***** *** the ****** / ********:

*******, *** ** *** few **** ************ *** Rekognition *** ** **** test ********, ****** *** light, *** ******* ****** straight ******* *** ****** but ***** **** *** not ******** ** ***:

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

*** ******** ****** **** is ************** *** * simplification. *********** **** ********* evaluation **** * ****** called"**** ******* *********" (***), ***** ******** ***********-**-******* on ****** **** ********* datasets ************** ** ****** **** bounding-box ******* (********* ****** "Intersection **** *****" ** IoU) *** ******* **** box ********* ******** * face. ** ****'* ***** library, **** **** ****** has *** **** (***** frame *** * ****, and ***** ** **** one). ***** ****, ** can *** * ****** heuristic *** ********: ** how **** ****** *** a **** ******** (******** / *****-***** = ********). This ** *** *******: in **********, ** **** not ******* *** ***** positives (* **** ** detected ***********, **** ** the **** ** ** the *******'* *****). **** happens ****** - ** we ***** **** *** heuristic. ** ****** ** more ******* **** ********** (lux, *****, ***) ******* having ** ***** ******** boxes *** ***** ***** of *** *******.

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

*** ******* **** ****** used *** *********** ***:

  1. ****** *** ***** ** Amazon **
  2. ***** *** *****
  3. ***** *********** (******** ** to *** ** ****, registering *****->***********)
  4. **** *** *****, ********* FPS

* **** ****** ******** for ************ ***** ** to ****** ****** ******** via*******->***********. **** ***** *** the ********** ******** ** a *****-**-***** *****, ***** our ***** ***** *** networking ******** **** *** FPS *********** ********. ** such, *** *** ***********, already **** ***, ***** be **** *****.

***** ** **** ************, this ** * ***** point ******* ***********, ** others ** *** ** a *******. *** *******, one *** *** ***-***** hardware *** ****** ******* performance. *** ***** ****** in **** ******* *** on ****** * ** Intel ** ******* (****-***) devices, ** **** *********** trade-off ** *********-*********. ************, Rekognition ** * ******* service, ***** *********** * model (**** ********'* ****-******-****) takes **** ************* *** setup. ** ***** ****-**** performance **** *********** ** unlikely (****: ** *** still ************* ** **** is ********), *** ***-****-**** applications, *********** ***** ** quite ******.

*******

*** * ****-**** ***** surveillance ***********,*********** ************ ** ***** **** but *** ***** ************ where **** ********** ***** processing ** ******, ** could ** *** **** affordable **** **-**** **********. For *******, *********** ******* for **-**** (*. ********) is $*.** *** * min ** **** ****** video ********. **** **** just **% ******** ** analyze (*.*., *,*** ******* in * *****), *** monthly **** ***** ** ~$50 (*,*** * **% x $*.**) **** *** the **** *********.

** *** ***** ****,****** *********** ****** ******* features********* ****** ***********, ****** analysis (***, ******, *******), object *** ********** *********, etc. **** **** ********* provider ****** ** *** have ** ***** ** likely ** ******** ** compete **** ***** ***********.

********

**** ** *** *** of ****'* *** ****** of **** ******** *****, starting ********** ********* ******,***** ****** ******* ***** 2 / ******** ** Test*** *** ****. ** are ******** * ******* of *** ***** ** the ******** ******. *** feedback, ******** ** *********, please *** ** *** comments.

Comments (8)

*** ***** ** *** interested ** *** ************,***** *** *** *** class **** ****:

deep learning webinar2

***** **** ** ********** fundamentals ** **** ** our ******** ****** ** provide ************* ******** *** education, ** ******** ** our ******* *****.

*** *** ******* **** running *** ********** **** IR *********** **? **** cameras ***** ****** ** IR *********** (** **** have *** ******) ** 1 ***

** *** *** ** any ** ******** *** we **** *** **** in ****** *****.

**** **** ***** **** integrate ** **** **** if **** *****'* *******.

** ***** *** *** you *** **** *** of **** *** ***** used ** *** ***** for **** ***** ** download *** *** **** their *** ********* *******?

* ***** **** ***** be * ****** ***** way ** ******* ********.

** ***** ******* *** dataset *** **** *** decided ***.

**** **** **** **% activity ** ******* (*.*., 4,380 ******* ** * month), *** ******* **** would ** ~$** (*,*** x **% * $*.**) just *** *** **** analytics.

****'* *** *********** *** "10%" ****** ***** ** this ***********? *,*** ******* itself ** **% ** total ******* ** * month.

Read this IPVM report for free.

This article is part of IPVM's 6,435 reports, 865 tests and is only available to members. To get a one-time preview of our work, enter your work email to access the full article.

Already a member? Login here | Join now

Related Reports

Facial Recognition 101 on Mar 18, 2020
Facial recognition interest, use and fear is increasing. This guide aims to...
Integrated IR Camera Shootout 2020 - Avigilon, Axis, Bosch, Dahua, Hanwha, Hikvision, Panasonic, Uniview, Vivotek on Jan 30, 2020
The best and worst cameras tested in this IPVM shootout showed major...
Bias In Facial Recognition Varies By Country, NIST Report Shows on Jul 15, 2020
While many argue that face recognition is inherently racist, results from one...
FLIR Screen-EST Screening Software Tested on Jun 30, 2020
In our FLIR A Series Test, the cameras' biggest drawback was their lack of...
IronYun AI Analytics Tested on Feb 17, 2020
Taiwan / US startup IronYun has raised tens of millions for its "mission to...
London Live Police Face Recognition Visited on Feb 13, 2020
London police have officially begun using live facial recognition in select...
FLIR A Series Temperature Screening Cameras Tested on Jun 04, 2020
FLIR is one of the biggest names in thermal and one of the most conservative....
Hikvision Fever Screening Thermal Solutions Examined on Apr 13, 2020
Hikvision is marketing "safer, faster, smarter" with their Fever Screening...
Hanwha Face Mask Detection Tested on Jul 01, 2020
Face mask detection or, more specifically lack-of-face-mask detection, is an...
New Axis M30 Cameras Tested on Mar 26, 2020
Axis has released a new generation of, for them, relatively low cost M30...
Clinton Public View Monitor (PVM) Mask Detection Tested on Jul 09, 2020
Face mask detection, or more specifically not wearing one, is expanding...
Video Surveillance History on May 06, 2020
The video surveillance market has changed significantly since 2000, going...
Face Shields Impact On Temperature Measurement And Mask Detection on Jul 27, 2020
First, the use of face masks, and now, plastic face shields are rising...
Directory of 201 "Fever" Camera Suppliers on Aug 04, 2020
This directory provides a list of "Fever" scanning thermal camera providers...
Hazardous & Explosion Proof Access Control Tutorial on Feb 27, 2020
Controlling access to hazardous environments requires equipment meeting...

Recent Reports

VSaaS Will Hurt Integrators on Aug 06, 2020
VSaaS will hurt integrators, there is no question about that. How much...
Dogs For Coronavirus Screening Examined on Aug 06, 2020
While thermal temperature screening is the surveillance industry's most...
ADT Slides Back, Disappointing Results, Poor Commercial Performance on Aug 06, 2020
While ADT had an incredible start to the week, driven by the Google...
AHJ / Authority Having Jurisdiction Tutorial on Aug 06, 2020
One of the most powerful yet often underappreciated characters in all of the...
SIA Coaches Sellers on NDAA 889B Blacklist Workarounds on Aug 05, 2020
Last month SIA demanded that NDAA 899B "must be delayed". Now that they have...
ADI Returns To Growth, Back To 'Pre-COVID Levels' on Aug 05, 2020
While ADI was hit hard in April, with revenue declining 21%, the company's...
Exposing Fever Tablet Suppliers and 40+ Relabelers on Aug 05, 2020
IPVM has found 40+ USA and EU companies relabeling fever tablets designed,...
Indian Government Restricts PRC Manufacturers From Public Projects on Aug 04, 2020
In a move that mirrors the U.S. government’s ban on Dahua and Hikvision...
Directory of 201 "Fever" Camera Suppliers on Aug 04, 2020
This directory provides a list of "Fever" scanning thermal camera providers...
Face Masks Increase Face Recognition Errors Says NIST on Aug 04, 2020
COVID-19 has led to widespread facemask use, which as IPVM testing has shown...
Dahua Loses Australian Medical Device Approval on Aug 04, 2020
Dahua has cancelled its medical device registration after "discussions" with...
Google Invests in ADT, ADT Stock Soars on Aug 03, 2020
Google has announced a $450 million investment in the Florida-based security...
US Startup Fever Inspect Examined on Aug 03, 2020
Undoubtedly late to fever cameras, this US company, Fever Inspect, led by a...
Motorola Solutions Acquires Pelco on Aug 03, 2020
Motorola Solutions has acquired Pelco, pledging to bring blue back and make...
False: Verkada: "If You Want To Remote View Your Cameras You Need To Punch Holes In Your Firewall" on Jul 31, 2020
Verkada falsely declared to “3,000+ customers”, “300 school districts”, and...