Using enebular For AI on the Edge

Using enebular For AI on the Edge

2018/10/15開催のAI IoTLTで使用したプレゼンテーション。講演者は、enebular開発部のFarhan Khanと山崎です。

The presentation deck used at AI IoTLT on Oct. 15, 2018. The main speaker was Farhan Khan, enebular development, and Wataru Yamazaki as the additional speaker.

40661cda330920ea2a854566ae19bbe9?s=128

wataru yamazaki (uhuru)

October 15, 2018
Tweet

Transcript

  1. 1.

    8 8 . 58 O A 5 2 2 .

    8 8 2 1 2 2 2 . EC K F I ,0 . . .
  2. 2.

    What? • Doing AI computation on Edge device instead of

    cloud. Why? • Quicker response. • High level of security (you don’t need to send data to cloud). • Low data transaction cost.
  3. 4.
  4. 5.

    : A A A A A A A A :

    A A . : A A . 2 A 2A :A : A A A A : A A A 2 A : A : 2 : A A 2A
  5. 6.

    3. Deploy AI model on the Edge device • Created

    enebular flow to deploy AI model from cloud to Jetson TX2 using enebular AI model deployment feature(alpha) • Once the model is deployed on the edge device, it will work even without internet connection. Uhuru engineers are working on this deployment feature and it will be available soon for all enebular users!
  6. 7.

     4 8 9 8 . • 0 52 /

    5 1 : • 333333
  7. 8.

        • . . sJ pn 21:

    21 • . . rl t S u v z C beka hk t C /: . • m k . . /1 . . ika() o fc _ 0:: . . 4 . . :
  8. 10.

    / /. / enebular 93$@ 105-0001 8"F;'CG 4-3-13  AD<

    4F +H 530-0005 +H2+H0&' !/ 3-2-4 !/  7F # ( 980-0021 .*?#(0IB' , 4-10-3 #(  2F S-234 )- 61 060-0031 &:E610 ,'& 1 78 1 > 6 = 5 61  3F 3 %54 @ WeWork 106-0032 8"F;'%54 1-4-5     16F Uhuru United Ltd. 2 Eastbourne Terrace, Paddington, London, W2 6LG . /