Getting Started

This page explains how to start-up the AI SDK.

AI SDK is build for specific board.
Its version varies depending on the supported board.

Supported Board	Supported AI SDK
RZ/V2L Evaluation Board Kit	RZ/V2L AI SDK v2.10
RZ/V2H Evaluation Board Kit	RZ/V2H AI SDK v3.00

Learn more about the RZ/V series.

Note Hereafter, we use following terminology.

• RZ/V AI SDK: Refers both RZ/V2L AI SDK and RZ/V2H AI SDK.
• EVK: Refers "Evaluation Board Kit".

Tutorial video Tutorial video is available here.

• For RZ/V2L EVK: RZ/V2L AI SDK v2.10
• For RZ/V2H EVK: RZ/V2H AI SDK v3.00

GUI environment AI SDK has a CUI environment and a GUI environment.
For the GUI environment, see AI Navigator Quick Start Guide.

Introduction

Getting Started Flow

Getting Started explains the instruction to run AI Applications.
Common instructions are explained in this document.
Board dependent instructions are explained in each EVK Getting Started page.

Software

Renesas provides following software.

Software	Provided on	Details
AI Applications	GitHub	This software is a set of following files that runs on the evaluation board kit. Application Source code (C++) Pre-build application binary Other necessary files to run the application Documentation (README.md) "Getting Started" page (this page) explains how to run the AI Applications.
RZ/V AI SDK	Renesas Website	This software is a development environment for AI Applications. Provided as a zip file. "Getting Started" page (this page) explains how to use RZ/V AI SDK to run the AI Applications.
RZ/V AI SDK Source Code	Renesas Website	This software is the source code of RZ/V AI SDK. Provided as a zip file. It includes Yocto Linux recipe. Details and instructions are explained in the following pages. How to Build RZ/V2L AI SDK How can I build RZ/V2L AI SDK Source Code? How to Build RZ/V2H AI SDK How can I build RZ/V2H AI SDK Source Code?

Step 1: Obtain an evaluation board

To start using RZ/V AI SDK, we need to get the board.
Renesas provides ideal board kit for evaluation.

Click the button below to get the board.
Board and Software Get the board and software for RZ/V series.

To see the details of the board, please see EVK Getting Started.
RZ/V2L EVK Getting Started What is included in EVK?
RZ/V2H EVK Getting Started What is included in EVK?

The RZ/V2L Evaluation Board Kit (EVK) is the ideal board kit for RZ/V2L evaluation.

Since MIPI camera module is included, you can start evaluating RZ/V2L immediately by building an environment.

Get RZ/V2L EVK

Following items are included in the RZ/V2L Evaluation Board Kit.

Equipment	Details
RZ/V2L Evaluation Board	Evaluation board itself.
MIPI Camera Module	Google Coral Camera. Note that the CMOS sensor (OV5645) in the camera is no longer available, and should not be used for mass production. Any software support provided is for evaluation purposes only.
MicroUSB to Serial Cable	For serial communication between PC and the board.

Step 2: Obtain necessary environment

1. Necessary Equipments

Please prepare the following equipments for your EVK.

For	Equipment	Details
RZ/V2L	RZ/V2L EVK	Evaluation Board Kit for RZ/V2L. Includes followings. MIPI Camera Module(Google Coral Camera) Note that the CMOS sensor (OV5645) in the camera is no longer available, and should not be used for mass production. Any software support provided is for evaluation purposes only. MicroUSB to Serial Cable for serial communication.
	AC Adapter	USB Power Delivery adapter for the board power supply.
	MicroHDMI Cable	Used to connect the HDMI Monitor and the board. RZ/V2L EVK has microHDMI port.
	USB Camera	Optional. AI Applications support USB camera input. Supported resolution: 640x480 Supported format: 'YUYV' (YUYV 4:2:2)
	Windows PC	Optional. Used as the serial communication console for eMMC Bootloader. Operating Environment : Windows 10
RZ/V2H	RZ/V2H EVK	Evaluation Board Kit for RZ/V2H.
	AC Adapter	USB Power Delivery adapter for the board power supply. 100W is required.
	HDMI Cable	Used to connect the HDMI Monitor and the board. RZ/V2H EVK has HDMI port.
	USB Camera	Since RZ/V2H EVK does not include camera module, this will be the standard camera input source. Supported resolution: 640x480 Supported format: 'YUYV' (YUYV 4:2:2) To use MIPI camera, please refer to e-CAM22_CURZH provided by e-con Systems. e-CAM22_CURZH camera supports FHD(1920x1080) resolution.
Common	USB Cable Type-C	Connect AC adapter and the board.
	HDMI Monitor	Used to display the graphics of the board.
	microSD card	Must have over 4GB capacity of blank space. Operating Environment: Transcend UHS-I microSD 300S 16GB
	Linux PC	Used for Setup microSD card and RZ/V AI SDK Setup. Operating Environment: Ubuntu 20.04
	SD card reader	Used for setting up microSD card.
	USB Hub	Used to connect USB Keyboard and USB Mouse to the board.
	USB Keyboard	Used to type strings on the terminal of board.
	USB Mouse	Used to operate the mouse on the screen of board.

Note USB camera has different supported resolution and format.
To check the specification of your USB camera, use v4l2-ctl command.

In addition to the RZ/V2L Evaluation Board Kit, please prepare the following equipments.

Note Serial to MicroUSB Cable and MIPI Camera Module (Google Coral camera) is included in RZ/V2L Evaluation Board Kit.

Equipment	Details
HDMI Monitor	Used to display the graphics.
Micro HDMI Cable	Used to connect the HDMI Monitor and the board.
AC adapter	Power supply to the board.
USB Cable Type-C	Connect AC adapter and the board.
microSD card	Must have over 4GB capacity of blank space. Operating Environment: Transcend UHS-I microSD 300S 16GB
Linux PC	Used for Setup microSD card and RZ/V2L AI SDK Setup. Operating Environment: Ubuntu 20.04
SD card reader	Used for setting up microSD card.
Windows PC	Optional. Used as the serial communication console for eMMC Bootloader. Operating Environment : Windows 10
USB Hub	Used to connect USB Keyboard and USB Mouse.
USB Keyboard	Used to type strings on the terminal of board.
USB Mouse	Used to operate the mouse on the screen of board.

2. Necessary Software

Please install following software on Linux PC.

• Docker
• git
• bmap-tools (For RZ/V2H only)

Step 3: Obtain RZ/V AI SDK

RZ/V AI SDK provides following packages for each supported board.

Name	Package	Details
RZ/V AI SDK	RTK0EF0*SJ.zip	Package used for AI development.
RZ/V AI SDK Source Code	RTK0EF0*SJ_linux-src.zip	Package used for Linux development. Please refer to following pages for more details on this package. How to Build RZ/V2L AI SDK How can I build RZ/V2L AI SDK Source Code? How to Build RZ/V2H AI SDK How can I build RZ/V2H AI SDK Source Code?

For this Getting Started, please download "RZ/V AI SDK" for your preferred device from the link below.
Board and Software Get the board and software for RZ/V series

Once downloaded, please check the Release Note included in the package.

Step 4: Extract RZ/V AI SDK package

This step explains how to extract the RZ/V AI SDK zip file.

1. On your Linux PC, make the working directory.

   mkdir -p ai_sdk_work

2. Register the working directory path to an environment variable.

   export WORK=<path to the working directory>/ai_sdk_work

3. Move to the working directory.

   cd ${WORK}

4. Extract RZ/V AI SDK zip file under the working directory.

   unzip <Path to the file>/RTK0EF0*.zip -d ${WORK}

5. Check the working directory to confirm the package contents.

   ls ${WORK}/

   • If the above command prints followings, the package is extracted correctly.

     ai_sdk_setup  board_setup  documents  references r11an0*.pdf

Step 5: Setup RZ/V AI SDK

This step explains how to setup the RZ/V AI SDK environment.

Note Make sure that you have installed Docker on your Linux PC.

1. On your Linux PC, move to the working directory.

   cd ${WORK}/ai_sdk_setup

2. Build docker image.
   
   • For RZ/V2L

     docker build -t rzv2l_ai_sdk_image --build-arg SDK="/opt/poky/3.1.21" --build-arg PRODUCT="V2L" .

   • For RZ/V2H

     docker build -t rzv2h_ai_sdk_image --build-arg SDK="/opt/poky/3.1.26" --build-arg PRODUCT="V2H" .

3. Create new directory to be mounted on Docker container.

   mkdir ${WORK}/ai_sdk_setup/data

4. Create docker container.
   Here, <rzv2l/rzv2h>_ai_sdk_container is a name of docker container, which can be changed by user.
   • For RZ/V2L

     docker run -it --name rzv2l_ai_sdk_container -v $(pwd)/data:/drp-ai_tvm/data rzv2l_ai_sdk_image

   • For RZ/V2H

     docker run -it --name rzv2h_ai_sdk_container -v $(pwd)/data:/drp-ai_tvm/data rzv2h_ai_sdk_image

   The local $(pwd)/data is mounted to /drp-ai_tvm/data on the Docker container by the above command option.
   For example, you can use this directory to copy files created on the Docker container to your local environment.
   
   


5. In docker container, run the following command to copy libtvm_runtime.so, which is the necessary file for the board, to host machine.
   

   cp /drp-ai_tvm/obj/build_runtime/${PRODUCT}/libtvm_runtime.so /drp-ai_tvm/data

6. To exit docker, run following commands.
   

   exit

7. To start the docker container again, run the following command.
   

   docker start -i <container_name, i.e., rzv2l_ai_sdk_container>

You have finished the RZ/V AI SDK environment setup.
You are ready to build the AI applications!

Step 6: Build RZ/V AI Application

This step explains how to build AI application.
To see the overview about AI Applications, please refer to About AI Applications page.
About AI Applications
What is AI Applications?
How is directory/file structured?
How can I download the previous version?

Note Following procedures are for users who would like to build the application from the source code.
If you would like to run the application as quickly as possible, you can skip this step and proceed to the next step (Step 7: Deploy AI Application) since pre-build application binary is provided.

Building instructions.

In AI Applications, there are two types of applications.

Application	GitHub Repository	Details	Example
<Application 1>	https://github.com/renesas-rz/rzv_ai_sdk https://github.com/Ignitarium-Renesas/rzv_ai_apps	These repositories provide the complete instruction in each application documentation.	R01_object_detection
<Application 2>	https://github.com/Ignitarium-Renesas/RZV2L_AiLibrary	This repository only contains AI Applications for RZ/V2L EVK. It does not provide instructions for RZ/V AI SDK. Please refer to the instruction provided in this page.	01_Head_count

• Option 1: For <Application 1>

  1. Check the README.md document provided in application directory and follow the instruction in the chapter called "Application: Build Stage" (or similar) to build the application.
     
     
     Example:
     In R01_object_detection application, follow the instruction here to generate the following application binary.

     • object_detection
     Note The git clone command shown in the README.md will download the latest source code and related files.
     
     To download the files of AI Applications v3.00, please specify the version tag by adding -b v3.00 when you running the git clone command as shown below.

     git clone -b v3.00 https://github.com/renesas-rz/rzv_ai_sdk.git

  
  After you generated the application binary, please proceed to Step 7




• Option 2: For <Application 2>

  1. Run (or start) docker container.
     
     
     Example:
     Run the following command to start the docker container created in Step 5.

     docker start -i rzv2l_ai_sdk_container

  
  
  2. Change the environment variable to use the cross compiler.

     source /opt/poky/3.1.21/environment-setup-aarch64-poky-linux

     Note This command needs to be called everytime user opened the new terminal.
  
  
  3. Move to the mounted point in docker container.
     

     cd /drp-ai_tvm/data

  
  
  4. Download the application source code.
     Here, we use RZV2L_AiLibrary repository 01_Head_count application as an example.
     

     git clone https://github.com/Ignitarium-Renesas/RZV2L_AiLibrary

  
  
  5. Move to the application directory.

     cd RZV2L_AiLibrary/01_Head_count/Head_count_cam

     Note The file configuration depends on each application. Please check the file configuration in the repository carefully.
  
  
  6. Build the application with make command.
     

     make

  
  
  7. Check that following application is generated in the RZV2L_AiLibrary/01_Head_count/
Head_count_cam/exe.
     • head_count_cam_app
  
  
  8. If you would like to build the other applications of <Application 1> (Option 1), please exit and restart the docker container.
  
  After you generated the application binary, please proceed to Step 7

Step 7: Deploy AI Application

Please see each EVK Getting Started Step 7 for detailed instructions.
RZ/V2L EVK Getting Started Step 7 How can I setup microSD card for eSD/eMMC Bootloader?
How can I deploy applications to the board?
How can I write eMMC Bootloader to the board?
How can I boot the board?

RZ/V2H EVK Getting Started Step 7 How can I setup microSD card for eSD Bootloader?
How can I deploy applications to the board?
How can I boot the board?

After Step 7, users should have completed followings.

• Setup miroSD card.
• Deploy the application on miroSD card.
• Boot the board.

Step 8: Run AI Application

Please see each EVK Getting Started Step 8 for detailed instructions.
RZ/V2L EVK Getting Started Step 8 How can I run applications?
How does application look on HDMI screen?

RZ/V2H EVK Getting Started Step 8 How can I run applications?
How does application look on HDMI screen?

After Step 8, users should have completed followings.

• Run the application.

Note To shutdown the board safely, please refer to followings.

• A3. Shutdown RZ/V2L EVK
• A4. Shutdown RZ/V2H EVK

This is the end of Getting Started.

You have experienced the AI Application development procedures.
Next step is to change the application to create your own AI Application.
You can use the applications listed in AI Applications to expand your ideas!

If you would like to customize Yocto Linux or develop your own board, please change and build the AI SDK Source Code based on your environment.
Please refer to How to Build AI SDK page shown below for more details.
How to Build RZ/V2L AI SDK How can I build RZ/V2L AI SDK Source Code?

How to Build RZ/V2H AI SDK How can I build RZ/V2H AI SDK Source Code?