Hailo-10H を Raspberry Pi 5 上で使用しています。
TAPPAS Core 5.3.1 (hailo-tappas-core_5.3.1_amd64.deb) を展開しましたが、
C++ example applications (apps/) が含まれていません。
Developer Zone では TAPPAS examples パッケージが見つからず、
GitHub の hailo-apps は Python 版のみで C++ apps がありません。
Raspberry Pi 5 (aarch64) 上でビルド可能な
TAPPAS example applications (YOLOv8 / segmentation / pose / multi-stream)
の提供をお願いできますでしょうか。
Welcome to the Hailo Community!
The hailo-apps repository also contains C++ examples:
For future posts, we kindly ask that you write your questions in English whenever possible. This helps both our team and other community members understand your issue and provide assistance more effectively. If needed, you can use a translation tool to translate your question before posting.
Thank you!
Thank you for your response.
Understood — I will write my future posts in English.
Thank you for pointing me to the C++ examples in the hailo-apps repository.
I would like to use the full C++ demo applications from the hailo-apps repository on my Raspberry Pi 5 with the Hailo-10H module.
Could you please provide access to the full TAPPAS SDK (including the C++ build environment, dependencies, and required models)?
I already have hailo-apps and tappas cloned, but the C++ demos cannot be built without the full SDK package.
Thank you!
Please follow the instructions to install all required packages. First follow the general Installation guide:
GitHub - hailo-apps - Installation
and then the instruction for each app e.g.
Thank you for the links.
However, the C++ applications in the GitHub repository require the full TAPPAS C++ SDK, including the compiled post‑processing libraries and GStreamer plugins.
The GitHub repository only contains partial sample code, not the actual runtime libraries needed to build and run the applications.
Could you please provide access to the full TAPPAS SDK package for Raspberry Pi 5 + Hailo‑10H?
Thank you.
Please send this now:
Thank you for your reply.
To clarify my situation:
I have already installed hailo-apps and followed the GitHub installation guide, but the C++ applications cannot be built or executed on Raspberry Pi 5 with the Hailo‑10H module because several required components are missing from the public repository.
Specifically, I need the following items, which are normally included in the full TAPPAS SDK:
TAPPAS C++ runtime libraries
GStreamer plugins for Hailo (e.g., hailo_postprocess)
Compiled post‑processing libraries (YOLO, segmentation, pose, depth, etc.)
Pre‑compiled HEF models required by the C++ apps
CMake toolchain files for building the C++ applications
Additional dependencies used by the C++ examples (OpenCV, GStreamer versions used by TAPPAS)
The GitHub repository contains only sample source code, not the actual runtime environment required to build and run the C++ applications.
Could you please provide access to the full TAPPAS SDK package for Raspberry Pi 5 + Hailo‑10H?
Thank you.
I think you may have missed the first part of the Raspberry Pi installation instructions.
sudo apt install dkms
sudo apt install hailo-h10-all
This installs most of the required dependencies.
Next, install hailo-apps :
# 1. Clone the repository
git clone https://github.com/hailo-ai/hailo-apps.git
cd hailo-apps
# 2. Run the automated installation script
sudo ./install.sh
After completing these steps, you should be able to build the C++ applications.
Thank you for the instructions.
I have already installed:
hailo-h10-all
hailo-apps
and I have already run install.sh
However, the C++ applications still cannot be built on Raspberry Pi 5 with the Hailo‑10H module.
The reason is that the GitHub installation only provides the Python-based environment.
The following components required for the C++ apps are missing:
TAPPAS C++ runtime libraries
GStreamer Hailo plugins (hailo_postprocess, etc.)
Compiled post‑processing libraries
Pre‑compiled HEF models
CMake toolchain files
TAPPAS‑specific OpenCV / GStreamer builds
These components are normally included in the full TAPPAS SDK, but they are not available in the public repository.
Could you please provide access to the full TAPPAS SDK package for Raspberry Pi 5 + Hailo‑10H?
Thank you.
Did you try building one of the C++ applications? What error messages do you get?
I set up a Raspberry Pi 5 with Raspberry Pi OS Debian GNU/Linux 13 (trixie) from scratch. I ran the commands mentioned above and then built the object detection application:
cd hailo-apps/hailo_apps/cpp/object_detection/
./build.sh
During the first build, I received an error instructing me to run:
git submodule update --init --recursive
After doing that, I ran the build again. The second build reported an error related to OpenSSL, which I resolved by installing the development package:
sudo apt install libssl-dev
After that, the build completed successfully.
I attached a USB camera and ran the application:
./build/x86_64/object_detection --input /dev/video8
Note: The x86_64 subdirectory name appears to be misleading, as the application is built and runs correctly on the Raspberry Pi. This is slightly different for each example application.
I also built and tested a few of the other examples, including oriented_object_detection and pose_estimation .
Here are a few snippets from the build output:
-I- Building x86_64
-- The C compiler identification is GNU 14.2.0
-- The CXX compiler identification is GNU 14.2.0
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Check for working C compiler: /usr/bin/cc - skipped
-- Detecting C compile features
-- Detecting C compile features - done
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Check for working CXX compiler: /usr/bin/c++ - skipped
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD - Success
-- Found Threads: TRUE
-- Found OpenCV: /usr (found version "4.10.0")
-- Found OpenCV includes: /usr/include/opencv4
-- Using CMake version 3.31.6
...
Install the project...
-- Install configuration: ""
-- Installing: /home/hailo/hailo-apps/hailo_apps/cpp/pose_estimation/build/x86_64/external/include/xtl/xbasic_fixed_string.hpp
...
[ 97%] Building CXX object CMakeFiles/pose_estimation.dir/yolov8pose_example.cpp.o
[ 97%] Building CXX object CMakeFiles/pose_estimation.dir/yolov8pose_postprocess.cpp.o
...
[ 97%] Linking CXX executable pose_estimation
Copy resources_config.yaml next to the executable
[ 97%] Built target pose_estimation
...
[100%] Built target curl-man
Thank you.
I tried building the C++ application as you suggested.
Here is the output from:
cd ~/hailo-apps/hailo_apps/cpp/object_detection
./build.sh
The build fails with the following error:
Could NOT find OpenSSL (missing: OPENSSL_CRYPTO_LIBRARY OPENSSL_INCLUDE_DIR)
CMake Error at FindPackageHandleStandardArgs.cmake:233 (message):
Could NOT find OpenSSL
Even if I install OpenSSL, the next build step fails because additional required components are missing (TAPPAS C++ runtime, GStreamer Hailo plugins, post‑processing libraries, HEF models, etc.).
It seems that the Raspberry Pi 5 + Hailo‑10H environment does not include the full TAPPAS C++ SDK, so the C++ applications cannot be built successfully.
Could you please provide access to the full TAPPAS SDK for Raspberry Pi 5 + Hailo‑10H?
Thank you.
These are installed by the various steps mentioned above.
Here are the steps, I executed on a fresh Raspberry Pi OS Debian GNU/Linux 13 (trixie). Change the /home/hailo path to your user path.
sudo apt update
sudo apt upgrade
sudo apt install dkms
sudo apt install libssl-dev
sudo apt install hailo-h10-all
cd /home/hailo
git clone https://github.com/hailo-ai/hailo-apps.git
cd /home/hailo/hailo-apps
sudo ./install.sh
git submodule update --init --recursive
cd /home/hailo/hailo-apps/hailo_apps/cpp/object_detection
./build.sh
Maybe access to some domains or sites are blocked for you. Make sure everything runs without errors.
There is no TAPPAS SDK for Raspberry Pi.
Thank you for the detailed instructions.
I will reorganize my development environment and clean up the previous directories to ensure a fresh setup.
For now, I will proceed with my own development workflow based on the information you provided.
Thank you again for your support.