ArmNN and TensorFlow Lite base mobilenet

Owned by edwin.chiu (Unlicensed)
Last updated: Jul 14, 2020
1 min read


This guide shows how we develop an application that classifies images using a TensorFlow Lite quantized Mobilenet V1 model. The guide also covers how we deploy the model using the open-source Arm NN SDK.
We will be able to use the knowledge from this guide to run our own models in our own applications on Arm Cortex CPUs.


Please follow below steps to get this demo.
1) Environment: SP7021 platform+Raspbian(Please refer to How to build and install SP7021 Linux image to run Raspbian on SD card)


2) Download mobilenet-quant source code, armnn and tensorflowlite library


    $ git clone https://github.com/sunplus-plus1/armnn_test.git


3) Extract armnn and tensorflowlite library


    $ cd armnn_test
    $ tar -xzf armnn-dist.tar.gz


4) Come in mobilenet-quant folder


    $ cd armnn-tflite-mobilenet-quant


5) Execute mobilenet-quant


    $ make test


6) Result



Input is malinois dog