Difference between revisions of "MISC-TN-011: Running an Azure-generated TensorFlow Lite model on Mito8M SoM using NXP eIQ"

From DAVE Developer's Wiki
Jump to: navigation, search
(→‎Introduction)
(→‎Introduction)
(14 intermediate revisions by 2 users not shown)
Line 1: Line 1:
 
{{InfoBoxTop}}
 
{{InfoBoxTop}}
 
{{AppliesToMito8M}}
 
{{AppliesToMito8M}}
−
{{AppliesTo MITO 8M TN}}
 
 
{{AppliesToMachineLearning}}
 
{{AppliesToMachineLearning}}
−
{{AppliesTo Machine Learning TN}}
 
 
{{InfoBoxBottom}}
 
{{InfoBoxBottom}}
 
{{WarningMessage|text=This technical note was validated against specific versions of hardware and software. What is described here may not work with other versions.}}
 
{{WarningMessage|text=This technical note was validated against specific versions of hardware and software. What is described here may not work with other versions.}}
Line 18: Line 16:
 
|March 2020
 
|March 2020
 
|First public release
 
|First public release
−
|-
 
−
|1.0.1
 
−
|April 2020
 
−
|Added TF model's graph
 
−
|-
 
−
|1.0.2
 
−
|April 2020
 
−
|Added TFL model's graph
 
 
|}
 
|}
−
 
 
==Introduction==
 
==Introduction==
−
In [[SBCX-TN-005: Using TensorFlow to implement a Deep Learning image classifier based on Azure Custom Vision-generated model|this Technical Note]] (TN for short), a simple image classifier was implemented on the [[:Category:AxelLite|Axel Lite SoM]].
+
In [[SBCX-TN-005: Using TensorFlow to implement a Deep Learning image classifier based on Azure Custom Vision-generated model|this Technical Note (SBCX-TN-005)]] (TN for short), a simple image classifier was implemented on the [[:Category:AxelLite|Axel Lite SoM]].
−
 
 
−
In [[MISC-TN-010: Using NXP eIQ Machine Learning Development Environment with Mito8M SoM|this other document]], it is illustrated how to run [https://www.nxp.com/design/software/development-software/eiq-ml-development-environment:EIQ NXP eIQ Machine Learning software] on i.MX8M-powered [[:Category:Mito8M|Mito8M SoM]].
 
  
−
This article combines the results shown in the TN's just mentioned. In other words, it describes how to run the same image classifier used in SBCX-TN-005 with the eIQ software stack. The outcome is an optimized C++ imaging classification application running on Mito8M SoM, which makes use of the eIQ software stack. In terms of hardware and software, the testbed used is the same described [[MISC-TN-010: Using NXP eIQ Machine Learning Development Environment with Mito8M SoM|here]].
+
In this [[MISC-TN-010: Using NXP eIQ Machine Learning Development Environment with Mito8M SoM|TN (MISC-TN-010)]], it is illustrated how to run NXP eIQ Machine Learning software on i.MX8M-powered [[:Category:Mito8M|Mito8M SoM]].
  
 +
This article combines the results shown in the TN's just mentioned. In other words, it describes how to run the same image classifier used in SBCX-TN-005 with the eIQ software stack. The outcome is an optimized imaging classification application written in C++ running on Mito8M SoM and that makes use of eIQ software stack.
 
==Workflow and resulting block diagram==
 
==Workflow and resulting block diagram==
−
The following picture shows the block diagram of the resulting application and part of the workflow used to build it.
 
−
 
−
 
−
[[File:MISC-TN-011-image-classifier.png|thumb|center|600px|Block diagram of the image classifier]]
 
−
 
−
 
−
First of all, the TensorFlow (TF) model generated with Microsoft Azure Custom Vision was converted into the TensorFlow Lite (TFL) format.
 
−
 
−
Then, a new C++ application was written, using the examples provided by TFL as starting points. After debugging this application on a host PC, it was migrated to the edge device (a Mito8M-powered platform, in our case) where it was natively built. The root file system for eIQ, in fact, provides the native C++ compiler as well.
 
−
 
−
For the sake of completeness, the following images show the graphs of the original TF model and the converted TFL model (click to enlarge).
 
−
 
−
{| class="wikitable" style="margin: auto;"
 
−
|+
 
−
!Original TF model's graph
 
−
!Converted TFL model's graph
 
−
|-
 
−
|[[File:Image-classifier-azure-model.pb.png|thumb|center|150px]]
 
−
|[[File:Image-classifier-azure-converted model.tflite.png|thumb|center|370px]]
 
−
|}
 
−
 
−
==Running the application==
 
−
The following block shows the execution of the classifier on the embedded platform:
 
−
<pre class="board-terminal">
 
−
root@mito8m:~/devel/image_classifier_eIQ# ./image_classifier_cv converted_model.tflite labels.txt testdata/red-apple1.jpg
 
−
Original image size: 600x600x3
 
−
Cropped image size: 600x600x3
 
−
Resized image size: 224x224x3
 
−
Input tensor index: 0
 
−
Input tensor name: Placeholder
 
−
Filling time: 25.3169 ms
 
−
Inference time: 276.121 ms
 
−
Total prediction time: 301.438 ms
 
−
Output tensor index: 406
 
−
Output tensor name: model_outputs
 
−
Top results:
 
−
0.997172  Red Apple
 
−
0.00214239 Green Apple
 
−
</pre>
 
−
 
−
The prediction time is cut by about 88% compared to [[SBCX-TN-005: Using TensorFlow to implement a Deep Learning image classifier based on Azure Custom Vision-generated model|this implementation]]. Of course, this is due to several factors. The most relevant ones are:
 
−
* i.MX8M is faster than i.MX6Q
 
−
* The application is written in C++ and not in Python
 
−
* The TF model was replaced with a TFL model, which is inherently more suited for ARM-based devices
 
−
* The middleware provided by NXP eIQ is optimized for their SoC's.
 

Revision as of 09:18, 25 March 2020

HOME SOMs SBCs ToloMEO Embedded Assistant Sale-tag.pngGET A QUOTE Customer-service.pngONLINE HELPDESK
Roadmap IoT Services ML/AI services Embedded Design Services
Info Box
DMI-Mito-top.png Applies to MITO 8M
NeuralNetwork.png Applies to Machine Learning
Warning-icon.png This technical note was validated against specific versions of hardware and software. What is described here may not work with other versions. Warning-icon.png

History[edit | edit source]

Version Date Notes
1.0.0 March 2020 First public release

Introduction[edit | edit source]

In this Technical Note (SBCX-TN-005) (TN for short), a simple image classifier was implemented on the Axel Lite SoM.

In this TN (MISC-TN-010), it is illustrated how to run NXP eIQ Machine Learning software on i.MX8M-powered Mito8M SoM.

This article combines the results shown in the TN's just mentioned. In other words, it describes how to run the same image classifier used in SBCX-TN-005 with the eIQ software stack. The outcome is an optimized imaging classification application written in C++ running on Mito8M SoM and that makes use of eIQ software stack.

Workflow and resulting block diagram[edit | edit source]

HOME SOMs SBCs ToloMEO Embedded Assistant Sale-tag.pngGET A QUOTE Customer-service.pngONLINE HELPDESK
Roadmap IoT Services ML/AI services Embedded Design Services
Retrieved from "https://wiki.dave.eu/index.php?title=MISC-TN-011:_Running_an_Azure-generated_TensorFlow_Lite_model_on_Mito8M_SoM_using_NXP_eIQ&oldid=9673"