Building Machine Learning Software with Reference Images for Yocto Project
Introduction
NXP eIQ software provides the enablement software for Machine Learning application optimized for i.MX SoCs.
eIQ offers Neural Network acceleration on NXP SoCs on the GPU or NPU through the usage of OpenVX as backend. Also, when executing inference on Cortex-A cores, NXP eIQ inference engines support multi-threaded execution.
eIQ is provided on a Yocto layer called meta-imx/meta-ml.
In this article, we will show how to integrate to the Toradex Reference Images for Yocto Project Software the following AI runtimes:
Toradex BSP Version | meta-ml version | AI Runtimes |
---|---|---|
Quarterly: 5.2.0 | Based on NXP BSP L5.4.70-2.3.0 Download documentation (requires login) | TensorFlow Lite v2.3.1 ONNX Runtime 1.1.2 OpenCV 4.4.0 |
The eIQ software based on NXP BSP L5.4.70-2.3.1
also offer support for the following AI Frameworks which we will add instructions soon:
- PyTorch 1.6.0
- Arm Compute Library 20.02.01
- Arm NN 20.01
All the AI Runtimes (except OpenCV, as documented on the i. MX Machine Learning User's Guide) provided by eIQ supports OpenVX (GPU/NPU) on its backend.
You can find more detailed information on the features of eIQ for each specific version on the i.MX Machine Learning User's Guide available on the NXP's Embedded Linux Documentation. See the version-specific information on the links in the table above.
You can also adapt the instructions to build on newer versions of BSP / meta-ml. However, we didn't test it at the time of this writing.
Converting Tensorflow Models to Tensorflow Lite
As stated in the Tensorflow Lite Documentation, the framework is:
TensorFlow Lite is a set of tools that enables on-device machine learning by helping developers run their models on mobile, embedded, and IoT devices.
In order to execute Tensorflow models with Tensorflow Lite, you need to use the Tensorflow Lite Converter. Please, note that the Tensorflow version used to design the model needs to match the Tensorflow Lite version.
Also, as stated on Tensorflow Lite documentation , not every model is directly convertible to Tensorflow Lite because some TF operators do not have a corresponding TFLite operator. However, in some situations, you can use a mix of Tensorflow and Tensorflow Lite ops. There is a list of Tensorflow ops that can be used with Tensorflow Lite by enabling the Select TensorFlow Ops feature. Please, see the Tensorflow Lite Documentation for more information about this feature and how to enable it.
Pre-Requisites
- One of the following Toradex SoM's:
- Verdin iMX8M Plus (CPU/NPU support)
- Apalis iMX8 (CPU/GPU support)
- Other i.MX-based SoMs may have support for CPU-only but are not tested. You can follow the article at your own risk.
- A compatible Carrier Board
- Optional: a USB webcam.
- Read the Build a Reference Image with Yocto Project article.
Adding eIQ recipes to Reference Images for Yocto Project
Clone the Toradex BSP repository
First, create a directory in your home named yocto-ml-build
and use repo to obtain the Toradex BSP on the version 5.2.0
, as explained in the section First-time Configuration of the Build a Reference Image with Yocto Project article:
info
To facilitate the comprehension of this article, we will create a directory in home called ~/yocto-ml-build
. You can, of course, use any name you want.
$ mkdir -p ~/yocto-ml-build/bsp-toradex && cd ~/yocto-ml-build/bsp-toradex
$ repo init -u https://git.toradex.com/toradex-manifest.git -b refs/tags/5.2.0 -m tdxref/default.xml
$ repo sync
info
At the time of this writing, we tested the building with the latest monthly release of the Toradex BSP Layers and Reference Images for Yocto Project Software, version 5.2.0
. You can use these instructions to build the meta-ml
with newer BSP versions; however, we didn't test it.
Getting eIQ
Git clone the meta-imx
repository to your ~/yocto-ml-build/
directory:
$ git clone -b zeus-5.4.70-2.3.1 git://source.codeaurora.org/external/imx/meta-imx ~/yocto-ml-build/meta-imx
Copying the Recipes to your environment
First, create a layer named meta-ml
, add it to your environment and remove the example recipe:
$ bitbake-layers create-layer ../layers/meta-ml
$ bitbake-layers add-layer ../layers/meta-ml
$ rm -rf ../layers/meta-ml/recipes-example
Copy the recipes from meta-imx
to your layer.
$ cp -r ~/yocto-ml-build/meta-imx/meta-ml/recipes-* ../layers/meta-ml/
$ cp -r ~/yocto-ml-build/meta-imx/meta-ml/classes/ ../layers/meta-ml/
$ cp -r ~/yocto-ml-build/meta-imx/meta-bsp/recipes-support/opencv ../layers/meta-ml/recipes-libraries/
This version of meta-ml
targets a version of Open-Embedded slightly different than the one of Toradex BSP version 5.2.0
. In order to build the tensorflow-lite python API with the BSP 5.2.0
, you need to make some adjustments on the python3native
recipe:
$ echo 'export _PYTHON_SYSCONFIGDATA_NAME="_sysconfigdata"' >> ../layers/openembedded-core/meta/classes/python3native.bbclass
$ sed -i 's/inherit siteinfo python3native/inherit siteinfo python3native setuptools3/g' ../layers/openembedded-core/meta/classes/meson.bbclass
Adding the recipes to your distribution
Add the meta-ml
recipes to your image:
$ echo 'IMAGE_INSTALL_append += "tensorflow-lite onnxruntime "' >> conf/local.conf
Add some image processing libraries to be able to execute additional image manipulations such as resize, crop, etc.:
$ echo 'IMAGE_INSTALL_append += "opencv python3-pillow adwaita-icon-theme "' >> conf/local.conf
In order to build the image a little bit faster, for now, we will remove the Qt packages. Keep it if you are planning to use Qt in your image.
$ echo 'IMAGE_INSTALL_remove += "packagegroup-tdx-qt5 wayland-qtdemo-launch-cinematicexperience "' >> conf/local.conf
Building
Build the tdx-reference-multimedia-image
image for your target SoM as explained on the Build a Reference Image with Yocto Project article.
info
In some situations of internet or server instability, the building may fail to clone some repository with an error similar to: do_fetch: Fetcher failure for URL:
. In most cases, this issue will be resolved by re-trying the building.
Flashing the image
To flash your image to the board, see the Quickstart Guide for your SoM.
Executing Demos
NXP provides an example for executing inference with and without GPU/NPU support. You can compare the inference time of each.
To execute it, cd
to the example's directory:
# cd /usr/bin/tensorflow-lite-2.3.1/examples/
This demo will take an arbitrary picture (grace_hopper.bmp
) as an input of an image classification neural network based on Mobilenet V1 (224x224 input size). See more information about this demo on the NXP's i.MX Machine Learning User's Guide
To execute the demo:
- NPU support (Verdin iMX8M Plus)
- GPU support (Apalis iMX8)
- CPU only (Apalis iMX8 and Verdin iMX8M Plus)
# USE_GPU_INFERENCE=0 ./label_image -m mobilenet_v1_1.0_224_quant.tflite -i grace_hopper.bmp -l labels.txt -a 1
# USE_GPU_INFERENCE=1 ./label_image -m mobilenet_v1_1.0_224_quant.tflite -i grace_hopper.bmp -l labels.txt -a 1
# ./label_image -m mobilenet_v1_1.0_224_quant.tflite -i grace_hopper.bmp -l labels.txt
Alternatively, you can run the same example using a Python implementation:
# python3 label_image.py
Please, be aware of the following limitation for the Python implementation, as stated on the NXP's i.MX Machine Learning User's Guide:
The TensorFlow Lite Python API does not contain functions for switching between execution on CPU and GPU/NPU hardware accelerator. By default, GPU/NPU hardware accelerator is used for hardware acceleration. The backend selection depends on the availability of the
libneuralnetworks.so
orlibneuralnetworks.so.1
in the/usr/lib directory
. If the library is found by the shared library search mechanism, then the GPU/NPU backend is used.
Therefore, if you want to evaluate the Python script without GPU/NPU support, rename the /usr/lib/libneuralnetworks.so.1
to /usr/lib/libneuralnetworks.so.1.backup
and execute it again.
info
As explained on the NXP's Application Note AN12964, the i.MX 8M Plus SoC requires an Warmup Time of about 7 seconds to initiate before delivering its expected high performance. You will observe this extra time when starting an application with NPU support.
Additional Resources
See the version-specific NXP's i.MX Machine Learning User's Guide for more information about eIQ enablement.