US Intelligence Agencies Research On Long-Range Biometrics Examined

By IPVM, Published Nov 11, 2020, 08:36am EST (Info+)

Identifying people at close range is increasingly accurate but what about identifying people at 100 or 500 meters or even 1 kilometer? This is far more difficult.

The US government is investing in developing these capabilities.

IPVM Image

IPVM researched and spoke with the office of the Director of National Intelligence's IARPA (Intelligence Advanced Research Projects Activity) group on what they want to achieve.

Executive *******

***** ** ********* *********** *** ************** at ******** *** ******* ************** ** ****** ********* **** **** at ********* ** ** ** * km **** ******** ******/***** ** ** to **°. ***** ***** ** ********* data, *******, ************, *** ******* *** teams ** ****** *** ********* **** the **** * *****. **** ****** it ** ******* ******** ********** **** face *** ****, *** *** **** to *** ********.

IPVM Image

**** ***** **** ***** *********' ***, where ********* ************ *** ***** **** about *** *******, **** **** ***** for *************, *** ***** ********, ***** more *********** ** *** ***** ******* and *** *****.

The ********* ** ******** *****:

******: *** ****** ******** ************* *** *** ***** *** has **** ******, ** ********** ******* *** ****** proposals *** *******.

IARPA ********

IPVM Image

******* * ******** ******** ** ********* States ****** ** *** ******** ** National ************ **** **** *** ******* on **** **** **** ********:

[******] ** ****-****/****-****** ******** ******** **** have *** ********* ** ******* *** nation **** ** ************ ************ *********.

***** **** **** ******* ******* *** “Bring[ing] *** **** ***** ** **** on *** ********” ** ****** ************* advances.

BRIAR ********

********* *********** *** ************** ** ******** and ***** (*****)** * ******* ** ****** ***********/************** to ******** ******* ** ********* ****** by************ ******** ******** ******** ******** (*****). **********'* **** *** ***********/************** ******** **** *** ** used **** * *****, ****** ****** on *** ** * *****-***** ********, and **** * ******** ** *** ground, ******* ***** ****. **** ******* a ***** ** ***-***** (***-*,*** ****) and ** ***** ** **°-**°.

IPVM Image

***** ***** ** ***-**-*** ****** **** takes *****-**** ****, ** ******** ******** and *** * *****-****** ******* **** can ** *** ** *** ****. They ***** **** **** **-** ************, so ******* **** *** ******** *** be **-********** **** * *****, ************ image ** **********.*** ******* ** ******* ** *******-***** technology.***** **** **** **** **** ********** multiple ***** ***** **** ** *** project **** ***** ************ ************* *** evaluating ********:

***** *********** ******** ******** ***** **** be ************ ** *** ********* ** the *******. ** ********, *** ***** program **** **** * ********* *** independent **** *** ********** **** **** will ******** *** ******** *** ******** deliverables ******* ******* *******, **********, *** goals. *** ***** ******* ******* **** oversee *** ******* *** ********** ************** and ************** ******* ***** *** **** stakeholders ******* *** **********.

***** **** **** **** ** *** have *** ******** ******* ** **** for *** ***** **********, *** *********** contributing ** ******* ******** *** ************ use-cases:

***** ******* *** ******** ********, ******* and ********** ** ****** *** ************ and ************ ********** *** ************ ********* (IC) *** *.*. **********. ***** **** not **** ** *********** *******. *******, the ***** ******* **** ****** ** contribute ** ******** ******** **** ******* counterterrorism, ********** ** ******** **************, ******** force **********, *** ****** ********.

******:*** ***** *** *** **** ********* ***** *** ***** ** ******** 10th *** *** ******* ***** ** funding. **** **** ****** *** ***** and ****** ** ** ***** *** confirmed *** ******** ** ****.

IARPA ******** ********

IPVM Image

***** ***** ** ******* *** ************* ****** **** *** ******* ********. They **** ** ****** *********** **** sets **** *** *******, **** **** consenting ***********, *** ******* ****** ***** from ******** ******* ** ********* ********* and ******, ********* **** ******. ***** is **** ********** ** ******* **** collection ******* *** ******** ******** **** sets **** ************ *** *******. ***** told **** **** ***** *** ******** how **** ******* **** ** *********** or ** **** ******. **** **** on ******* ******** *** *** “******** to ******** ********” ** **** ** the ********* ** “*********, ******, *** permanent”. **** *********** *** ******** **** be * ****** ** ****, ****, and ***** **********:

*** *****, ********, ************, ** ***** aspects ** * ***** **** (********* face)

IPVM Image

Challenges ** ********

**********, ***** *******, *****/*****, *** **** of **** *** *** ************ ************** ** ****** *** ********. At * ********, ********** ****** ****** to ** *********.********* ******* ********* *** *** ******** and ***** ******* ****** ******** ** prior ******.

IPVM Image

***** ******* **** ****** *** ***** of *********** **** *** ** ****. This ** **** *********** *** ********** that *** ******* *******, **** **** recognition.********* ******** ***** *** ***** ** obstacles ** ******** **********, ****** **** are *** ******** *** ***** ******** angles ******** ** ****** ** **** camera *********. *******, ***** ******** ******** data ** ****** *** **** *******, unrealistic, ** *********. **** **** ** creating *** ******* ***** *** ******** to ******** ****.

3 ****** **** * *****

*** ******* ** ****** **** **** 3 ******, ******** **** ****** ***-***** ** medium-pitch, **** ****** ****-**** ** ****-***** or ****-*****, *** ******* ********* *** two *** ****** **** *********.

IPVM Image

Phase *:** *** ***** ** ******, ***** expects * **% **** *** **% face ************** ******** **** *% ***** identification ** * *** ****** *********. That ***** **% ** **% ** correct ******* **** ***** ****** *** 99% ** ****** *** *** *** on *** *** ****** ********* **** not *********** ***** ******. **** **** on ***** ****** ***** **** *-**** (1700 ****) ** ****** **** *°- **°******** ****** ** ********* ********** (*.*. turbulence *** ******* ********). ***** ******* video ********** ** ****** **** ****-**** with $**,*** ** ********.

Phase *: ** *** ****** ** ******, ***** expects * **% **** *** **% face **** ************** **** **** *% false ************** ** * *** ****** watchlist. **** ***** **% ** **% of ******* ******* **** ***** ****** and **% ** ****** *** *** not ** *** *** ****** ********* will *** *********** ***** ******. **** plan ** ***** ****** ***** **** 1-1000m (**** ****) ** ****** **** 0°- **° ******** ****** ** ********* ********** and ****-**** ********** **** $**,*** ** hardware.

Phase *: ** *** ***** ** ******, ***** expects * **% **** *** **% face **** ************** **** **** .*% false ************** ** * *,*** ****** watchlist. **** ***** **% ** **% of ******* ******* **** ***** ****** and **.*% ** ****** *** *** not ** *** *,*** ****** ********* will *** *********** ***** ******. **** plan ** ***** ****** ***** **** 1-1000m (**** ****) **** *°- **°******** ****** ** ********* ********** **** real-time *** **** ********** (***** ******** requirements *** **** *********). ***** *** not ******* ******** ***** *** **** phase.

IPVM Image

Previous *********

IPVM Image

***** ******** *************** *** **** *********** ** * past ******* **** **** **** ** emulate **** *****. ******** ******* *** ********* **** ** ******* **** *********** and***** ***** ** *** ************* **** *** ***** **** ******* as ********** ********* ******* ***** ************. Top **** *********** ********** ****** *********** **** ********** **%****** *** ******* ** **** ** 99.8% ** *** **** ****** ******* this ****. **** *********** *** ******** Vision********* ******** *** ** ******** ********* **** **** ** ******** * lack ** **** *** ******* *** been ******* **** *********** ****, *** that ** ******** ***** **** ****** accuracy ********* ******* ** **** ***********. He ********** **** ** ***** *** no *** ***** ******* **** *********** could ******* *** ******** ** *** achieved ***** *** ******* ******* ******** from ***** ** ****/***** **** **********.

Companies ********

***** *** ***** ******* *** *** officially *******,******* ****** **** ************* ****** ** *********** *** **** listed ************, ** ****** *** ******** them ** ********* ** "****" ****** the *******.

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

  • ** *********, ******** ***** *********** *** ********** mitigationIPVM Image
  • ***** ***********, **** **** **** *** ** face *********** ** ********* ** * km
  • ****** ********, *** **** **** **** ****** on ********* ******* *** ******
  • ******* *******, ***** ****** ** **** ********* re-id **********
  • ****, *** ***** ****** ********** *** capturing ******-***** ******** ****.

Comments (8)

IARPA told IPVM they do not have any specific mission in mind for the BRIAR technology

What mission is there other than target identification and apprehension or destruction? They aren't picking specific people out from a watch list with a drone at 1000m to hand out cookies. Unless cookies is some kind of nickname for a Hellfire missile ;-)

Agree
Disagree
Informative
Unhelpful
Funny: 5

target identification and apprehension or destruction

Those are two very different things and the technology could be used in a lot of different ways. One use of this technology could be saving money on access control by allowing face recognition to be integrated with existing cameras that may be too high or at too steep an angle for face recognition. Another use could be surveilling a city from drones that are high enough to be ignored. And, a third use could be turning drones with explosive devices into targetted missiles that search a city for a target.

Agree
Disagree
Informative: 1
Unhelpful
Funny

intruders at great distances before they become a threat

Agree
Disagree
Informative
Unhelpful
Funny

695 Campbell Technology Parkway, Campbell CA

Agree
Disagree
Informative
Unhelpful
Funny

IPVM Image

1984 was supposed to be fiction, not a prediction.

Agree: 1
Disagree
Informative
Unhelpful
Funny

So... I wonder what would have been said if China or Chinese manufacturers did this.

Ok. That aside, when this is complete and someone with power tripping tendencies takes control of this "tool", what would stop them from say targeting a certain demographic for whatever reason.

You see I mention this because I am on several Surveillance forums and I constantly see the bashing of Chinese manufacturers who are even in any way involved in a venture like this. They are called human rights violators because someone uses their "tool" improperly... so I ask what is the difference here.

Agree
Disagree
Informative
Unhelpful
Funny

They are called human rights violators because someone uses their "tool" improperly... so I ask what is the difference here.

This is a good point, but there are a few differences.

1. Uyghur (or any ethnicity) detection analytics are inherently discriminatory. Creating them while the PRC government is targetting this group is not creating a tool that is used improperly, it is creating a tool that can only be used improperly while knowledge of what the tool is being used for.

2. Manufacturers know that the PRC government is using surveillance technology in Xinjiang to monitor the Uyghur minority. Many are selling these tools with the knowledge that they will be used for human rights violations.

3. This project hasn't started yet. There has been no technology created so there haven't been any uses to examine. IPVM will follow the project to see how it develops, and if successful, how the technology is used.

That being said, recognition and identification analytics are capable of being misused, and it is important for manufacturers to ensure they can trust end-users not to abuse their technology. There are currently no rules specifically governing how this technology could be used, including as Undisclosed Integrator #1 implied, self-targeting weapons.

Agree
Disagree
Informative
Unhelpful
Funny

Update: The BRIAR BAA has been posted and teams can apply by February 10th for the initial round of funding. IPVM will follow the story and follow up as teams are confirmed and progress is made.

Agree
Disagree
Informative
Unhelpful
Funny
Login to read this IPVM report.
Why do I need to log in?
IPVM conducts reporting, tutorials and software funded by subscriber's payments enabling us to offer the most independent, accurate and in-depth information.
Loading Related Reports