MIVisionX

MIVisionX toolkit is a set of comprehensive computer vision and machine intelligence libraries, utilities, and applications bundled into a single toolkit. AMD MIVisionX delivers highly optimized conformant open-source implementation of the <a href="https://www.khronos.org/openvx/" target="_blank">Khronos OpenVX™</a> and OpenVX™ Extensions along with Convolution Neural Net Model Compiler & Optimizer supporting <a href="https://onnx.ai/" target="_blank">ONNX</a>, and <a href="https://www.khronos.org/nnef" target="_blank">Khronos NNEF™</a> exchange formats. The toolkit allows for rapid prototyping and deployment of optimized computer vision and machine learning inference workloads on a wide range of computer hardware, including small embedded x86 CPUs, APUs, discrete GPUs, and heterogeneous servers.

Latest release

AMD OpenVX™

AMD OpenVX™ is a highly optimized conformant open source implementation of the <a href="https://www.khronos.org/registry/OpenVX/specs/1.3/html/OpenVX_Specification_1_3.html" target="_blank">Khronos OpenVX™ 1.3</a> computer vision specification. It allows for rapid prototyping as well as fast execution on a wide range of computer hardware, including small embedded x86 CPUs and large workstation discrete GPUs.

<a href="https://www.khronos.org/registry/OpenVX/specs/1.0.1/html/index.html" target="_blank">Khronos OpenVX™ 1.0.1</a> conformant implementation is available in MIVisionX Lite

AMD OpenVX™ Extensions

The OpenVX framework provides a mechanism to add new vision functionality to OpenVX by vendors. This project has below listed OpenVX modules and utilities to extend amd_openvx, which contains the AMD OpenVX™ Core Engine.

amd_loomsl: AMD Loom stitching library for live 360 degree video applications
amd_media: AMD media extension module is for encode and decode applications
amd_migraphx: AMD MIGraphX extension integrates the <a href="https://github.com/ROCmSoftwarePlatform/AMDMIGraphX#amd-migraphx" target="_blank"> AMD's MIGraphx </a> into an OpenVX graph. This extension allows developers to combine the vision funcions in OpenVX with the MIGraphX and build an end-to-end application for inference.
amd_nn: OpenVX neural network module
amd_opencv: OpenVX module that implements a mechanism to access OpenCV functionality as OpenVX kernels
amd_rpp: OpenVX extension providing an interface to some of the ROCm Performance Primitives (RPP) functions. This extension enables rocAL to perform image augmentation.
amd_winml: AMD WinML extension will allow developers to import a pre-trained ONNX model into an OpenVX graph and add hundreds of different pre & post processing vision / generic / user-defined functions, available in OpenVX and OpenCV interop, to the input and output of the neural net model. This extension aims to help developers to build an end to end application for inference.

Applications

MIVisionX has several applications built on top of OpenVX modules. These applications can serve as excellent prototypes and samples for developers to build upon.

Neural network model compiler and optimizer

Neural net model compiler and optimizer converts pre-trained neural net models to MIVisionX runtime code for optimized inference.

rocAL

The ROCm Augmentation Library - rocAL is designed to efficiently decode and process images and videos from a variety of storage formats and modify them through a processing graph programmable by the user.

rocAL is now available as an independent module at https://github.com/ROCm/rocAL. rocAL is deprecated in MIVisionX.

Toolkit

MIVisionX Toolkit is a comprehensive set of helpful tools for neural net creation, development, training, and deployment. The Toolkit provides useful tools to design, develop, quantize, prune, retrain, and infer your neural network work in any framework. The Toolkit has been designed to help you deploy your work on any AMD or 3rd party hardware, from embedded to servers.

MIVisionX toolkit provides tools for accomplishing your tasks throughout the whole neural net life-cycle, from creating a model to deploying them for your target platforms.

Utilities

loom_shell: an interpreter to prototype 360 degree video stitching applications using a script
mv_deploy: consists of a model-compiler and necessary header/.cpp files which are required to run inference for a specific NeuralNet model
RunCL: command-line utility to build, execute, and debug OpenCL programs
RunVX: command-line utility to execute OpenVX graph described in GDF text file

Prerequisites

Hardware

CPU: AMD64
GPU: AMD Radeon™ Graphics [optional]
APU: AMD Radeon™ Mobile/Embedded [optional]

[!IMPORTANT] Some modules in MIVisionX can be built for CPU ONLY. To take advantage of Advanced Features And Modules we recommend using AMD GPUs or AMD APUs.

Operating System

Linux

Ubuntu - 20.04 / 22.04
CentOS - 7
RedHat - 8 / 9
SLES - 15-SP5

Windows

Windows 10 / 11

macOS

macOS - Ventura 13 / Sonoma 14

Installation instructions

Linux

The installation process uses the following steps:

ROCm-supported hardware install verification
Install ROCm 6.1.0 or later with amdgpu-install with --usecase=rocm
Use either Package install or Source install as described below.

Package install

Install MIVisionX runtime, development, and test packages.

Runtime package - mivisionx only provides the dynamic libraries and executables
Development package - mivisionx-dev/mivisionx-devel provides the libraries, executables, header files, and samples
Test package - mivisionx-test provides ctest to verify installation

Ubuntu

sudo apt-get install mivisionx mivisionx-dev mivisionx-test

CentOS / RedHat

sudo yum install mivisionx mivisionx-devel mivisionx-test

SLES

sudo zypper install mivisionx mivisionx-devel mivisionx-test

[!IMPORTANT]

Package install supports HIP backend

Package install requires OpenCV V4.6 manual install

CentOS/RedHat/SLES requires FFMPEG Dev package manual install

Source install

Prerequisites setup script

For your convenience, we provide the setup script, MIVisionX-setup.py, which installs all required dependencies.

python MIVisionX-setup.py --directory [setup directory - optional (default:~/)]
                          --opencv    [OpenCV Version - optional (default:4.6.0)]
                          --ffmpeg    [FFMPEG Installation - optional (default:ON) [options:ON/OFF]]
                          --amd_rpp   [MIVisionX VX RPP Dependency Install - optional (default:ON) [options:ON/OFF]]
                          --neural_net[MIVisionX Neural Net Dependency Install - optional (default:ON) [options:ON/OFF]]
                          --inference [MIVisionX Inference Dependency Install - optional (default:ON) [options:ON/OFF]]
                          --developer [Setup Developer Options - optional (default:OFF) [options:ON/OFF]]
                          --reinstall [Remove previous setup and reinstall (default:OFF)[options:ON/OFF]]
                          --backend   [MIVisionX Dependency Backend - optional (default:HIP) [options:HIP/OCL/CPU]]
                          --rocm_path [ROCm Installation Path - optional (default:/opt/rocm ROCm Installation Required)]

[!NOTE]

Install ROCm before running the setup script

This script only needs to be executed once

ROCm upgrade requires the setup script rerun

Using MIVisionX-setup.py

Clone MIVisionX git repository

git clone https://github.com/ROCm/MIVisionX.git

[!IMPORTANT] MIVisionX has support for two GPU backends: OPENCL and HIP

Instructions for building MIVisionX with the HIP GPU backend (default backend):
- run the setup script to install all the dependencies required by the HIP GPU backend:
```
cd MIVisionX
python MIVisionX-setup.py
```
- run the below commands to build MIVisionX with the HIP GPU backend:
```
mkdir build-hip
cd build-hip
cmake ../
make -j8
sudo make install
```
- run tests - test option instructions
```
make test
```
Instructions for building MIVisionX with OPENCL GPU backend

Windows

Windows SDK
Visual Studio 2019 or later
Install the latest AMD drivers
Install OpenCL SDK
Install OpenCV 4.6.0
- Set OpenCV_DIR environment variable to OpenCV/build folder
- Add %OpenCV_DIR%\x64\vc14\bin or %OpenCV_DIR%\x64\vc15\bin to your PATH

Using Visual Studio

Use MIVisionX.sln to build for x64 platform

[!IMPORTANT] Some modules in MIVisionX are only supported on Linux

macOS

macOS build instructions

[!IMPORTANT] macOS only supports MIVisionX CPU backend

Verify installation

Linux / macOS

The installer will copy
- Executables into /opt/rocm/bin
- Libraries into /opt/rocm/lib
- Header files into /opt/rocm/include/mivisionx
- Apps, & Samples folder into /opt/rocm/share/mivisionx
- Documents folder into /opt/rocm/share/doc/mivisionx
- Model Compiler, and Toolkit folder into /opt/rocm/libexec/mivisionx

Verify with sample application

Canny Edge Detection

export PATH=$PATH:/opt/rocm/bin
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/opt/rocm/lib
runvx /opt/rocm/share/mivisionx/samples/gdf/canny.gdf

[!NOTE]

More samples are available here

For macOS use export DYLD_LIBRARY_PATH=$DYLD_LIBRARY_PATH:/opt/rocm/lib

Verify with mivisionx-test package

Test package will install ctest module to test MIVisionX. Follow below steps to test packge install

mkdir mivisionx-test && cd mivisionx-test
cmake /opt/rocm/share/mivisionx/test/
ctest -VV

Windows

MIVisionX.sln builds the libraries & executables in the folder MIVisionX/x64

Use RunVX to test the build

./runvx.exe ADD_PATH_TO/MIVisionX/samples/gdf/skintonedetect.gdf

Docker

MIVisionX provides developers with docker images for Ubuntu 20.04 / 22.04. Using docker images developers can quickly prototype and build applications without having to be locked into a single system setup or lose valuable time figuring out the dependencies of the underlying software.

Docker files to build MIVisionX containers and suggested workflow are available

MIVisionX docker

Documentation

Run the steps below to build documentation locally.

sphinx documentation

cd docs
pip3 install -r sphinx/requirements.txt
python3 -m sphinx -T -E -b html -d _build/doctrees -D language=en . _build/html

Doxygen

doxygen .Doxyfile

Technical support

Please email mivisionx.support@amd.com for questions, and feedback on MIVisionX.

Please submit your feature requests, and bug reports on the GitHub issues page.

Release notes

Latest release version

Changelog

Review all notable changes with the latest release

Tested configurations

Windows 10 / 11
Linux distribution
- Ubuntu - 20.04 / 22.04
- CentOS - 7
- RHEL - 8 / 9
- SLES - 15-SP5
ROCm: rocm-core - 6.1.0.60100
RPP - 1.5.0.60100
miopen-hip - 3.1.0.60100
migraphx - 2.9.0.60100
OpenCV - 4.6.0
FFMPEG - n4.4.2
Dependencies for all the above packages
MIVisionX Setup Script -