RZ/V AI SDK Overview

AI Software Development Kit (AI SDK) is an AI application development environment for Evaluation Board Kit of RZ/V series.

Since it contains Yocto Linux with bootloader, Linux Kernel, Cross Compiler and a complete set of libraries for DRP-AI and graphics, users can develop AI Applications easily and quickly with AI SDK.
In the AI Applications, USB/MIPI camera can be used with Video for Linux 2 (V4L2) API.

AI SDK is specifically made for the supported Evaluation Board Kit with fixed Linux components. For those of you who would like to customize Linux environment, i.e., change memory map or develop your own board, you need to build the Yocto Linux source code.
Yocto Linux source code is provided as AI SDK Source Code.

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?


AI SDK has a CUI enviroment and a GUI environment.
Each has the following functions.

Functions	CUI env.	GUI env. (e2 studio)
Install AI SDK	✔	✔
Build AI Applications	✔	✔*1
Build AI SDK	✔	-
Transfer Learning	-	✔*1

*1: There is a limit to the number of AI Applications supported.


For more information, see each AI SDK download page.
Board and Software Get the board and software for RZ/V series.

Key Features

Supported Devices

• RZ/V2L Evaluation Board Kit : RZ/V2L AI SDK Specification
• RZ/V2H Evaluation Board Kit : RZ/V2H AI SDK Specification

File Contents

AI SDK provides following files.

Package	Items		Details	RZ/V2L AI SDK	RZ/V2H AI SDK
AI SDK	Release Note		PDF documentation that contains RZ/V AI SDK release information.	Contains list of files provided in RZ/V AI SDK.
	board_setup	Bootloader	Files to boot the board.	Provided as individual files. Supports eSD/eMMC Bootloader.	Included in WIC format image. Supports eSD Bootloader.
		Linux Kernel Files	Pre-build binary files for RZ/V Linux Kernel.
		Root filesystem	Filesystem for the RZ/V Linux.
	ai_sdk_setup	AI SDK Installer	Installer for AI application development environment on Docker.	Provided as a Dockerfile that clones DRP-AI TVM .
		Cross Compiler	Compiler for Linux application on RZ/V Linux, which is used by AI SDK Installer.	Provided as cross compiler installer.
		DRP-AI Translator Installer	Installer for DRP-AI Translator, which is used by AI SDK Installer.	Provided as an installer taken from DRP-AI Translator .	Provided as an installer taken from DRP-AI Translator i8 .
	documents	Related Documents	PDF documentation that contains RZ/V AI SDK related information.	Includes DRP-AI Translator User's Manual and list of Linux license information.
	references	Reference information	Files that is refered by Linux license information documentation.	Includes linux manifest file, etc..
AI SDK Source Code	README		Release information of AI SDK Source Code	Provided as a txt file.
	Yocto Linux Recipe		Linux source code of AI SDK.	Provided as a tar file.
	OSS Source Code		Source code of Open Source Software pakcages used to build AI SDK.	Provided as a 7z file.	Not provided in v4.00.

To check more details, please see each Release Note/README provided in the package.

RZ/V2L AI SDK Specification

Target Version: 2.10

Overview

RZ/V2L AI SDK configuration is as follows.

• DRP-AI TVM v1.1.1 [ GitHub ]
• RZ/V2L Linux Package v3.0.4 [ Renesas Web ]
  • Video for Linux 2 (V4L2)
  • Other device drivers
    (See RZ/V2L Linux Package Release Note.)
• Graphics Library Evaluation Version v1.1.0 ^*2 [ Renesas Web ]
• Video Codec Library Evaluation Version v1.1.0 ^*2 (H.264) [ Renesas Web ]
• RZ/V2L DRP-AI Support Package v7.40 [ Renesas Web ]
  • DRP-AI Driver
  • OpenCV 4.1.0
• Other OSS packages
  • Tesseract 3.05.01 (tesseract, tesseract-lang)

For more information on RZ/V2L AI SDK, see the page below.
RZ/V2L AI SDK Configuration See Linux Components information.


For more information on AI SDK Source Code, see How to Build AI SDK page.
How to Build RZ/V2L AI SDK How can I build RZ/V2L AI SDK Source Code?

*2: Note that Graphics Library and Video Codecs Library Evaluation Version have a limitation that system stops running after around 10 hours.
If you would like to use unristricted version, please build AI SDK with Graphics Library Unristrictd Version and Video Codecs Library Unristricted Version according to How to Build RZ/V2L AI SDK.

Memory Map

The memory map of RZ/V2L AI SDK is as follows.

RZ/V2H AI SDK Specification

Target Version: 4.00

Overview

RZ/V2H AI SDK configuration is as follows.

• DRP-AI TVM v2.2.1 [ GitHub ]
• RZ/V2H Linux BSP [ Configuration ]
  • DRP-AI Driver v1.00 [ GitHub ]
  • e-CAM22_CURZH camera driver (MIPI) ^*3
  • Video for Linux 2 (V4L2) including the support for USB camera
  • Other device drivers [ List ] [ Manual ]
• Graphics Library Evaluation Version v1.2.1 ^*4
• Video Codec Library v3.0.0 ^*5
• OpenCV Accelerator v1.00 (supporting OpenCV 4.1.0) [ GitHub ]
• Other OSS packages
  • Tesseract v3.05.01 (tesseract, tesseract-lang)

For more information on RZ/V2H AI SDK, see the page below.
RZ/V2H AI SDK Configuration See Linux Components information.
See Drivers list.
See Limitations.


For more information on AI SDK Source Code, see How to Build AI SDK page.
How to Build RZ/V2H AI SDK How can I build RZ/V2H AI SDK Source Code?

*3: To use e-CAM22_CURZH camera, see e-con Systems.
*4: Note that Graphics Library Evaluation Version has a limitation that system stops running after around 10 hours.
If you would like to use unrestricted version, please build AI SDK with Graphics Library Unrestricted Version according to How to Build RZ/V2H AI SDK.
*5: Video Codec Library support only H.264 Enc in AI SDK v4.00.
If you would like to use H.264 Dec and H.265 Enc/Dec function,
please refer to Appendix 3: Prepare Video Codec Library for H.264 Enc/Dec and H.265 Enc/Dec function.

Memory Map

The memory map of RZ/V2H AI SDK is as follows.

Back to Home >