Healthy Skepticism for Deep Learning Is Prudent

Author: John Honovich, Published on Jul 26, 2017

The hype for deep learning in video surveillance is accelerating.

Between the race to the bottom and dearth of a 'next big thing', certainly pent up demand exists.

But now is the time for a healthy skepticism about when and what deep learning can do. In this note, we examine the hype and real issues that professionals need to consider.

*** **** *** **** ******** ** ***** ************ ** ************.

******* *** **** ** *** ****** *** ****** ** * 'next *** *****', ********* **** ** ****** ******.

*** *** ** *** **** *** * ******* ********** ***** when *** **** **** ******** *** **. ** **** ****, we ******* *** **** *** **** ****** **** ************* **** to ********.

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

'Revolution' ****

**** *****,*** ******** **** ********** ** "* ********** ** ***** *********", *********:

********************** **** ***** **** ********** ******* **.* ******* ******** ** ************, ***** ************ ******* ***** ******** ** ******* ** *******.

************* **** ***** ***** ***, **** ***** ***** ** *** even *** *** **.* ******* ** ***** ******** *** *** mentioned.

*** **.*% ****** **********.

Irony ******

*** ***** ** *** *** ***** ** *** ******* ******** of *** **** ******** *****:

*** ***** ********* ******** ** ***** ******* *** *** ************ damage ** ********* ** * ****** **previously ********* *** **** and delivering too little in the past. [emphasis added]

*** ****** "***** *** ****** ******** *** **** *** ********* to ****** **" ** ********* *********** ****.

Robots

**** ** ****** ***, '******* *******' ******** ******* *********** ************ *** ****** *** ****** *** **** *** "**** ** robots **** ****** ********** *******". ** *******:

** **** ** ** ** ******** ********* ***** **** **** a **** **** ***

** ********, ** *** **** ******** ***** *** *****. ** are ** ** '******** ********* *****' **** ****** **** ****. Unfortunately, *** *******, ************** **** ****** * ****** ****** ***** ****.

Timing ******

** ** **** ** *** **** ********** *** ************ **** become ********** ** ********* ****** **** ******. *** ******* **** eventually *** *** *** ***** ******.

*** ****** ****** ** ***********.

***** *** / ***** ** ***** ****** ********. *** ********* are ****. *********** *** ******** *****.

Deep ******** **********

**** *** * *** ********* ******* *** * ******* ********** of **** ******** ** *********:

  • ******* ************: **** ******** **** ** **** ** ******* ***** ************ applications. ***** **** **** ****** **** *** ***** **** **** poorly ** ***** *******. ** ** ****** ******* **** *** performance **** **** ************* ****** ************. "**** ********" ** ****** to ** ******** ** ********* *** ***** ** **********.
  • ***** *********** ********: ***** **** ******** ************* ****** **** ********, *** *********** will ** ************* ******** ** **** ** ** *** ************ **. *** '**** *****' ** ******* *** ** ***** applications *** ******** ** ********* *******, ******* *** ***** (*.*., people **** *********, ****** ***********), ** ** ******* **** *********** problems **** ** *****. **** ** *** ******* ** *** easily, ******* ** ******* *** **** ******** ************ *** ** re-trained ** ****** ***** *********.
  • ******* ****** ***********: ****** ******* ***** **** ******** **** ******* **** * few *********, *** ******** *** *********** ** **** ******** ***** surveillance ******* ** ***** ** **** ************* ***** ** *** engineering ***** (***** ** ** ***** ****** *******), ********* ********* and **** ********* ** ****** *** ******* ****** ** ***** on. **** ** ****** ** ****** ** *********** *********** ********* across ******* **** ** *** ** **** *** '**** ********' on *** '***'.
  • ******** ***** ******: ****** ***********,*** **** ******** ******** **** *********** ******** ***** ******.

Early *******

**** *** *** *****, ** ** ********* ** ****** ********* that **** ** ***** ***** **** *** **** ******** ** video ************.

* *** ** **** ** ***** ***** ** ***** ********* demos. *** *******, ** *** ****** *** **** **** *****, one **** ******** ******** *** ******* *** *** ******** / performance ** ***** ******. ***** * ****** ** ** ** watching **, *** ***** ****** **** *** **** ****** **** on ******* **. ** ****** ***, **** **** ****** ******* in * **** ** *** *** ******. ***** ***** ****** will **** ********** **** *** *** *** * *** ** deep ******** ** **** *****.

Healthy **********

**** **** ********, ** **** **** * ****** ** ***** for **** ******** ** ****** (**** *********** *** ******** ****) for ** ** **** * *** ********** ** ***** ************. Hyping ** *****, ********** ***** *** ***-**** ******* **** ***** analytic ******** ** *** *****, ***** ********* ******** **** ********'* reputation *** *** ********** ********* ****** **** ** **.

Comments (11)

*** ** *** *** ************ ******** ****** **** **** ********* the ****** *** ******** ** '**** ********' ** ************/******** ************??

*.*. *** ** ********* ** ********?

************ ******** ** * **** ********.

*** *******, * ***** ****** '********' ** *** ******* ** a '********' ***** ** ******** *** **** ****** ***** *** it.

* ** *** **** *** ****** *** ** **** *** around *** **** ********* ******* *** ******** (*.*., ** **** learning ***** ************ '*******' ** **/**'* *******?). @*****, *** **** you ****.

*.*. *** ** ********* ** ********?

***, ******** ** **** ***** ** ***** *********** ********** (*.*.:*.*** ********* **** ******** / *** *********), ******'* ******** *********, ***. ****** **** *** ********* *** defined ** "**********" ************.

**** ** *** ********* ******* ******* ****** ***** (*********, ***.) applied ** ********** *******. **** ******** ** *********, ** ***** so ***, *** **** ******* ******* ** *** ********** *******, doing * ****** *** ** *********** **** ** "*****" (*******, boat, ***, ***) ** *** *****. ***** *** ******* ********* applications ** ****, *** ***** *** *** *********** ***** ******** unique (*** ****** ****** ********* *********** ******).

*** ***** ****** * ******** **** ******** ******* **************, *** it ***** *** ** * **** ****** ******. **** ** very ******* ********** ** ********* * ***** ***** ********** ***. You ***** ** **, *** ***** *** ** **** **** to ***** * ********** *** **** ******* ***** ***** * competing *** ******* ********** **** ****** ** ***. ******** ******** used ** ***** ***'* *** *** **********, *** ***** *** some ***** ******** ** **** ***** ******** *** ********** *** (many *** ********* *** ***-**********/******** ********).

*** ***** ****** * **** ****, ****-******** ******* *** ******* to ******* **** ** ******* ********* *** ******** ******** ** a ****** ** *******, *** *** ***** *** ***** ******* to *** ** (** *** *** * ******* ***). ****, if ******* **** *** ***** **** ******, *** *** *** tagging ** ***** ***, *** ***** *** ******* * ******* fee (**** ** ******** *** ****** ********** *** *** *****/*********** by ***).

**** ******** ** ***** ******* ** ** **** ***** ** so **** ******** **** ** ** ******** **** ******** ******* will **** ***** ************ ******* ** **** ******** **** ****** the ******** ********.

***** *** ***** *** **** ****** ******** ** ******** ** in ***-*** ********/************ ***** ** *** ** ************ **** **** learning ***** ********. *** *******, *** ***** ** **** ** patent * *** *****-****** ******* ****-******** ******* ** * *** that ** ********* ******** *** ** * *** ****** ********* a ******, ** ******** * ********** ******** ******* ********.

**** ** * ***** *******... ************:

"*** ****** ****** ********* *********** ******"

** ** ******* ************* ** '**** ********' - *** **** me ** ** * ** ***** - ** ******* ** me ** ** **** ** *** ******** ******* ***** ** pretty ****** '**/****' ********* ***** ************.

**** *** *********** *** '**** ********' ********** ** * ******** increase ** ************* ***** *** *****.

*** ****** ***** **/**** ********** *** ** ******, *** **** operational ************ **** ******* **** ***** '**** ********' ** '****' things **** **** *** ******** ******* *** ************* ***** ******** to ***** **** ****.

** * *****?

*** **/**** ******* ** *** * *** *******, *** ** is **** *** * ******* *******. * ** *** **** if *** *** ***** ******* ***** *******/************* ** ***-***** ************, which ***** ****** **** *** *** *** **** ** **** the **/**** *******.

**** ******** ** *** ** (********** ************), ** ** **** about ******* ****** **** **** *** ********, *** **** ***** pattern *********/***********, ************ ** ***** ***** ****** ***** *** ** as **** ** ***** **** **** *** ****.

*** *** ***** * *** ** ****** ***** ***** ** cancer, *** **** **** *** **** ** *** ***** ** know *** ** ********* ** ********* ****** ********* ********.

******* - *'* *** ** ******** *** *** ********* ** not ***** ******....

***** *** * ******** ******* ***** ****** ** **** (***** vs. *** ****) **** ***** *** *** *** ****** **** an *********, ** **** ********. ** **** ** **** ** hardware. **** ** ********* * *** ****, *** **** ***** waves ******* **** ******** ***** **********.

****** ******** ** **** ******** *******, **** ********, **** ******* this ****** ** ********* *** ******** ** ******** ******* ** not ********.

*********, ** ************* ** **** *** *** *** ****** **** Learning **********, *** ****** **** **** ** **** ** ******** like ********** *** ******* ** * *** **** ** *********** to *** *********.

*****...

"* ** *** **** ** *** *** ***** ******* ***** patents/patentability ** ***-***** ************, ***** ***** ****** **** *** *** you **** ** **** *** **/**** *******."

**, ******** * ** ** ********* - * ***'* ***** that ********* *********** ****** ** ********** ******. *** '*****-******' **** set *** ** ********/****** ** ****** *********** ******. **'* **** simple ****.

**** ***** *** ******** ** ********* (* *****) ** *** ability ** **** '*****' **** *** ******* ***** **** **** fast ****** ** ****** * ********* *** **** *** ****** that *****.

** **** ****** *** ** *** ***** *****, *** ********* in *** ******** **** ** ** ********* *** **** **** they **** **** *** ***** ******* - ******** **** **** word **** ********** **** *******.

********* ************* ***** *********** ** ******** ** *********** ***** ****** on * **** **** ********** ** **** ** ******* **** really **** ******.... *** ** **** ******** ** ****-***** (** VCA ***) ** * ******* ****** **** ***** *** *** capabilities ** ******** ******* ******** ******** ********, *** **** ******** has *** ********* ** ***** ****** ** *** **** - like *** ********* ********** *** ****** **** **** *** ************ to ******** ******** *****.

** ** * ******* *****, *** ** *** ** *** early ******.

***** **** ***** ********* ******** ********* (** ******** ******) *** in ***** ******** ****** **************. **** ******** ***** ******* ***-****** approaches ** *** ** * ******** ****** ** ********** ***** from ***-***** (***** **** *****, **** ******** ********, ******* ****-******* coupled **** ******** ****** **** ** ********* ***** *******, ***.)

** ***-***** ********** *** ****** ******** **** ** **** ** object**, **** ***** ****** *** ** ***** ** *** ** that ** * *******/******/**** ******** ******.

** **** ****, *** ***** ** *** ******** ***** ** analytics. ****** *****, ** *** ****** ** ***** ******* **** an ******** ********* ******, ** ******* **** *******, *******, *** expensive. ***** ****** * ***** * **** ** ******* *** embedded *******, ******* ********* *********** ** *** ******/************* **** **** than *** ** (*** ** ***** ******* **** ****-****** **** of ***** ******** ** ***** ****** ******** ****** ********* ***** get *****).

******** *** ** ********* ***** *** *** ******** **** * major ******, *** *** ***** **** **** ****** ********** (*** marketing/sales) *** *** ********** ***** **** ***** ***** *** **** figured *** **** ** ****** ** ** ***** **.

* ***** ************* ************* ** ********* ** **** ** *** ** ******** *** security ********.

*** *** ****** ********* ** **** *** *** **** ** research ** *** *********.****://***.**********.**/.

*********, **** ** **** ** * *********** **** **** ** actual ***, ************ *** ******** ************. **** ***** ****** ** DNN-based ********* ****** *** **** ********** *** **********, ** ********* below.

* ***'*/***'* ******* ** *** "**** * *****" ******** ******* I ** *** **** ****** ***** *** ********** ******* ********* work ** * *** *****.

*** ******** ************, *** *********** ****** ***** ********* ** **** form ** ******* **********, ** **** * ***** ****** (**** likely * ******) ***** ***** ********* ** *** ******. **** could ** ** **** *** ***** ******** ** "***", ** to ****** ***** ********* ********** (******** * ****** ******** ** be ********* **. **** ***-***********).

***** *** * *** ****** **** **** ** *********/**** ********* to ** ** *********** ****** *** ************:

  1. ******** ********* *** ********* *** **** ******/************/****** ******** *** ******* around **. ************* ******* ***** ********** **-*****, **** ****** *****. The ******* ******* ****** ********, ***** **** ** ***** ******** cannot ** ****** ****** ****** **** ***** ** ****** ** attackers ** *** ****** ** * ****** *** *&*/******* ********. Not ****** ** ** **********, *** ** **** **** ****** a ****** ****** *** *** ******* ** **** ******* **** of ***** ********.

  2. ****** ********* **** ******* ******* ** ** ****** ****** ******** giving ** * **************. ******** ******* **** *** ******* ** generally **** ****** ***** ****** *******. ****, ******** ********* ** image ** *************** *********, ** ** ***** ** ***-****** ****** out ** *** ***** *** **** ******* ******** **** **** some **** ********. **** ** ********* * ******* **: **** for **** ******->******* *** ***** ****/*****/****** ***** ********->******* *** "*********** motion"->then ******* *** *** **** ******* *** *** *** **************. This ***** *** *** **** ** ** **** **** **** inject ******** ****** ****** **** *** ****** ** ****** **.

  3. ***** ****** ******* ***********, ** ********** ** ****** *** *** would ****** *** **** ** *******, ****** **** **** ******* with *** **** ******* (********, *** ****** ** ** ******** more ****** ** *** ****** ** ***** ****** ** *******/********* labeled **** **** ********), *** ******* *** ****.

*********, ** **** * ******* ****** ********-***** *** ********* ******, you ***** **** ** ********* **** **** ** ******* **** that ******** ********** *******, **** ******** ****** *** ***** ** motion (***** ***** ******* ******** ********** ** **** ******). ** the ****** ***** ******** ******** *** ********* ****** *** **** 1 ******, ** ***** ******** ** *****.

***** ** ******** ***********, ** ** **** **** ** **** which, ** ***, ********* *** ** *** (* ***** **** is **** * ****** ********** *** ******** *** ***** ***********-******** cameras). ******** *** * ********** ****** **** **** ****** ***** require **** ****** ** ****** ********* *** * **** ****** motivated ********. ** ******* ***** *** *** ***** ** ************ for ***** ***** *********, ** *** **** ** *** ****** today, *** ******** *** *********** *** *** ***** ** ****, those ********** *** ******* *** **** **** *** *****-****** *****-*****-***-*********-***** armed ******.

*** **** ** * *********** ********, *** ********* **** ******* would ****** *** ******** ******** *** ** ** ****.

* ***** ***** ******** ** ******** ******* *** ***** ***-******* existing ****** **** ** ******** *** ******** ***** *** ****** layer *** ***** *****. ***** **** ** *** ****** ******** and ******** ***** *** ******.

** * **** ** *****: ****** **** *** **** ** optimize *** **** ** *******. ****** ****** **** *** ******** to ****** *********** ************; ****** ******* ** ***** *** ***** distribution *** *** ********* ** *********** ********. ********* *** *** resistant ** *********** ********.

******* ************** ** ***** **** ** ********* ** **** ******** and ******* ***** ******** ****'* **** ** ***** **** ** optimise *********** ******, ********* ******** ** ******** ** ********* *** working *** **** ****. (**** *** ***** ** **** *****: https://arxiv.org/pdf/1412.6572.pdf)

** **** ********* *********** **** ****** **** ***-*******, ****-***** ****** model ********** ***** ***** ********** ** * ******* **** **** some ***** ** * *****.

** ** ****, **** (********** *********** ********) ** ** ****** area ** ******** *** ** ***** * **** ******** ********* and *** ******* ** ****** ******.

Login to read this IPVM report.
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.

Related Reports

Beware Amazon Go Store Hype (Tested) on Nov 13, 2018
IPVM's trip to and testing of Amazon Go's San Francisco store shows a number of significant operational and economic issues that undermine the...
Genetec Privacy Protector Tested on Nov 12, 2018
Genetec has built Kiwi Security's Privacy Protector into Security Center, an analytic which anonymizes individuals in cameras' fields of view...
Axis: "No One Wants To Buy A Camera" on Nov 09, 2018
Axis has, in its own description, made a bold declaration: The industry is changing so rapidly that the following statement might seem bold but...
Kogniz Silicon Valley AI Startup Profile on Nov 07, 2018
Kogniz is a Silicon Valley company that aims to bring AI analytics to security and surveillance, centering on their own smart cameras: We spoke...
Avigilon Opens Up Analytics And Cameras on Nov 06, 2018
Avigilon is opening up. The company historically famous for advocating its own end-to-end solutions and making it harder for 3rd parties to...
Huawei Admits AI "Bubble" on Oct 16, 2018
A fascinating article from the Chinese government's Global Times: Huawei’s AI ambition to reshape industries. While the Global Times talks about...
Dahua Face Recognition Camera Tested on Oct 15, 2018
Dahua has been one of the industry's most vocal proponents of the value that AI creates: As part of this, Dahua has released a facial...
IACP 2018 Police Show Final Report on Oct 08, 2018
IPVM went to Orlando to cover the 2018 IACP conference, the country's largest police show (about as big as ASIS), examining the 700+...
Genetec Takes Aim At 'Untrustworthy' 'Foreign Government-Owned Vendors' on Sep 24, 2018
Genetec is taking aim at 'untrustworthy' 'foreign government-owned vendors'. This is not a new theme for Genetec as nearly 2 years ago, Genetec...
Avigilon Announces AI-Powered H5 Camera Development on Sep 19, 2018
Avigilon will be showcasing "next-generation AI" at next week's ASIS GSX. In an atypical move, the company is not actually releasing these...

Most Recent Industry Reports

Throughtek P2P/Cloud Solution Profile on Nov 15, 2018
Many IoT manufacturers either do not have the capabilities or the interest to develop their own cloud management software for their devices....
ASIS Offering Custom Research For Manufacturers on Nov 15, 2018
Manufacturers often want to know what industry people think about trends and, in particular, the segments and product they offer.  ASIS and its...
ISC East 2018 Mini-Show Report on Nov 15, 2018
ISC East, by its own admission, is not a national or international show, billed as the "Largest Annual Northeast U.S. Security...
Hikvision Silent on "Bad Architectural Practices" Cybersecurity Report on Nov 14, 2018
A 'significant vulnerability was found in Hikvision cameras' by VDOO, a startup cybersecurity specialist. Hikvision has fixed the specific...
French Government Threatens School with $1.7M Fine For “Excessive Video Surveillance” on Nov 14, 2018
The French government has notified a high-profile Paris coding academy that it risks a fine of up to 1.5 million euros (about $1.7m) if it...
Integrator Credit Card Alternative Divvy on Nov 13, 2018
Most security integrators are small businesses but large enough that they have various employees that need to be able to expense various charges as...
Directory of Video Intercoms on Nov 13, 2018
Video Intercoms, also known as Video Door-Phones or Video Entry Systems, have been growing in the past decade as more and more IP camera...
Beware Amazon Go Store Hype (Tested) on Nov 13, 2018
IPVM's trip to and testing of Amazon Go's San Francisco store shows a number of significant operational and economic issues that undermine the...
Magos Radar Company Profile on Nov 12, 2018
Magos America General Manager Yaron Zussman admits when he first came across Magos, he asked himself: "What's innovative about radar?" Be that as...
Genetec Privacy Protector Tested on Nov 12, 2018
Genetec has built Kiwi Security's Privacy Protector into Security Center, an analytic which anonymizes individuals in cameras' fields of view...

The world's leading video surveillance information source, IPVM provides the best reporting, testing and training for 10,000+ members globally. Dedicated to independent and objective information, we uniquely refuse any and all advertisements, sponsorship and consulting from manufacturers.

About | FAQ | Contact