Neurala Deep Learning Platform Profile

Background *******

******* *** ******* ** 2006 ** * ****** University ***********, ************ ******* [**** no ****** *********] (***) ******* **** Versace [**** ** ****** available] (COO) *** ******* ************* [**** no ****** *********] (***). *** company ******* *** **** seed *******, *** ******, and ***** ****** * total ** ~$*** ** develop **** ****** ******* frameworks ******** *** *** in ***-***** ***'*. ***** aim *** ** ****** devices **** ******, ******** cameras, *** ******** *********** with ********-****** ************ ********** unavailable ** ***** ********** due ** ****, ****, or ***** **********.

*** ******* ******* ~** people ** *** ****** headquarters, *** ******** ** them ** ***********/*&* *********.

Edge ******** - *** **************

*******'* ********** ****** **** devices ** ** *********** trained ** ****-**** ** new ****** *******, ********* that ***** ********* ******* more ******** *********** ********** resources, *** ********** **** delays (*****). ******* ****** ** this ********** ** ** LDNN, ** ******** **** Neural *******.

**** *******'* **** ******** on * ***** ******, a **** *** ******* video ** *** ******, highlight ** ****** ** the *****, *** ***** the ****** ** ********* future *********** ** **** object.

**** ********** ****** **** other "**** ********" ********* which ********* **** ******* understanding ** ******* ******* objects, *** *** ****** to ***** ** ******** new ****** ***** ** appearance ******* *** *** of *********** *********.

Nvidia ************

*** ****-***** ******** *** been*********** ** ******'* ****** GPU. ******* ** **** part ********'* ********* *******, ******** ** **** foster ******* ********* **** utilize ****** ******** *** deep ******** ************.

Motorola ****

******** ** ***** *******'* investors, ********* ** *********** sum *********'* $*** ****** *** **** ** *** partnership ******* *** *** companies. ********'* ***** ******** ******* *** ******** ** **** advantage ** **/*** ************, enabling *******'* ******** ** run ****** ***** *******.

Bodycam ************ ** ******* ****

** ********* *******'* ********** in ********, *** ******* *** be *********** ******* ** recognize ******* ** ******* of ********, ******** *** officer ******** ** ********* when ** ******* ** appearance. ******** ********* ******** the ******* "*******" *** camera ****** ** *** person/object ** ** **** for ********. * ***** image *** ** ****, or * ***** **** can ** **** (**** as **** ******** ******** footage, ** **** *******), though **** ******** ******* are ******** ** ***** a ****** ** ****** (still ** *****) ******* of * ****** "*******" style *****. *** ******* then ********* *** ****** locally, ********* ** * period ** * *** minutes, ** ****** ** updated ******* ** *******, which ** *** **** recognize ** **** ***** from **** ***** *******.

*** ********* ****** *** camera *** **** ** used ** **** ********** on ******* ** ******* encountered ** ** ******* during *** ****** ** their *****. **** *** provide *********** **** ********** such ** *** ***** an ******* ****** *********** walking *****, ****** ** people, **********, ***** ********, street *******, ***. **** data *** ** **** to ******** ****** ******, or ***** ***** ****** within *** ************.

Surveillance ************ ** ******* ****

***** ************ ******* *** typically ** *** **** any **** ********* ********, and *** ********* ** a ********, ***** *** **** ********** power, *** **** *** edge-learning ****** * ****** itself ** *******. *******, embedded ********* ***** ****** low-cost ***********, *** **** far ***** *********, ***** Neurala's ********** ** ********* could ******* ************ ******* to ***** ****** ** *********'* **** ******** ********,******* ********* * ****** card **** ******. **** ** ******-***** recorders **** **** *********, the ******* ** ** dynamic ******** ** * way **** ******* ******* requirements *** ** ********** to ******** **** ***********.

***** ******* **** ****** edge-based ********* ************ *** object **************, * ****** of ************* **** ** Avigilon, ****, ******, *********, and ******, **** ******* begun ********* ******** **** edge-based *********, ****** *******'* offering ** **** ****** of ***** ******** *** the ************ ******.

Upcoming ******* ** **** ******** ****

******* ****** **** **** soon ******* * ******** update **** ****** ****** training **** ** *** device ** ** ******** and ****** ** ***** devices ****** *** **** organization. **** ***** ***** an ***** *** ******* their **** ****** ** the ********** ** * person ** ***** **** custom ******** **** **** other ********. 

Looking *** ******** ******** ********

*** ******* ****** **** would **** ** ******* with ******** ******** ************* to ******* *******'* ********** in ***** ********. **** would ****** ****** *** appropriate *** ** *** device, *** **** ********* Neurala's ********. *** *** requirement ***** **** **** manufacturers ***** **** ** create *** ******** ******* for ******** ******* ***** they ****** ** ****** the **********.

Challenges *** ******* ** ********

***** *******'* ********** *** be ********** *** ************* to ******** ******* ************, adding even ***-*** *** ************ can ******** ****** ****** by **'* ** *******, which *** ** * significant ******** ** ***** for **** ******* *** recorders. ************, ***** ************* **** been ******** ** **** on *** ******** *** core ******** **** ** analytics, ********** ******* ** create ***** *** ******** as * ***** ** differentiation. *******, ***-**** ******* may **** *******'* ********** *** offer **** * *** to ******* ***-******* ******** without ****** ** ****** heavily ** ***** *** R&D.