AI Model Deployment in Edge Computer

JavaScript isn't enabled in your browser, so this file can't be opened. Enable and reload.

Looking forward to hear from your 360 degree view

Email *

Deployment of “trained CNN model in X86 machine ” is having the following items as part of its deployment.

a. Ubuntu 18.04 OS is used in deployment machine which is X86

b. Python is not required in deployment machine which is X86

c. Tensorflow is not required in deployment machine which is X86.

d. FPGA (via PCI add on ) is not required in deployment machine which is X86

e. GPU (via PCI add on ) is not required in deployment machine which is X86

f. Item (b) and item (c) aree always true

g. Item (e) may be true some time

Other:

Clear selection

Deployment of “trained CNN model in Android Phone” is having the following items as part of its deployment .

a.NDK is ( in Android Studio ) used to build Inference engine which is developed in C and C++

b.Inference Engine is not using GPU in phone

c.Inference Engine is not using DSP in phone

d.Inference Engine is capable of using updated trained cNN model from Ubuntu machine (X86) via Wi-Fi

e.Inference Engine is used to collect data from USER ( int hsi case LKG student )

f.Inference Engine is showing inference output in display

Other:

Clear selection

Deployment of Tensorflow AI NN / CNN model or Python based model in IoT edge is required to optimize and quantize trained NN/CNN model. This happens because

CNN model in Python is not ready to deployed in IoT edge

CNN model in Tensorflow is not ready to deployed in IoT edge

Trained CNN model might have floating point weights

Trained CNN model might have too many Neurons

IoT edge technology is very different from Deep learning based inference technology

All the above are true

Other:

Clear selection

Provide your thought process to find a way to use DLtrain to solve issues in daily workflow in managing AI deployment workload.
*

In case, if you need a discussion with Expert, your email can be shared.

First Name *

Last Name *

In Your Deep Learning, how do you plan to do Inference ?
*

Submit

Clear form

Never submit passwords through Google Forms.

This form was created inside of IoT consultant.

Does this form look suspicious? Report

Forms