Introduction
The VisionPack AI Application Zoo applications are bundled as part of Torizon for Maivin and can be launched directly using systemctl commands described in this article.
The VisionPack AI Application Zoo sample applications are built using the VisionPack UI library (VPKUI) and come in two flavours: appnamegl and appnamegl_headless. The GUI for the applications uses OpenGL on the 3D GPU to provide high-resolution graphics in real-time. The _headless variants stream the GUI over HTTPS using H.264 which is helpful on a device such as Maivin which does not have an output display connector, the graphics can be streamed over Ethernet or WiFi.
Requirements
- Quick Start Guide
- Maivin AI Vision Starter Kit
- Torizon for Maivin 6.5.0.2 or newer
- Refer to the Updating Torizon for Maivin article for instructions on updating your Maivin.
- SSH into the Maivin
- If you SSH into the Maivin, and output the system information ( uname -rv ), it should look as follows:
torizon@verdin-imx8mp-06976874:~$ uname -rv
5.15.129-6.5.0.2-devel+git.6f8fd49366db #1-TorizonCore-maivin SMP PREEMPT Mon Mar 18 18:55:40 UTC 2024
Usage Instructions
The applications are bundled on Maivin along with a systemctl service definition to handle launching of the applications with the correct parameters. The Maivin bundled versions of the applications are headless and are built to capture camera from the Maivin Camera topic of the EdgeFirst Middleware.
For example, the Face Detection example can be run using this command:
sudo systemctl start facedetectui
List of Applications
The following is the list of demo applications.
- Detection Application
-
sudo systemctl start detectionui
-
- Face Detection Application
-
sudo systemctl start facedetectui
-
- Face Blurring Application
-
sudo systemctl start faceblurui
-
- Head Pose Application
-
sudo systemctl start headposeui
-
- Body Pose Application
-
sudo systemctl start bodyposeui
-
- People Segmentation Application
-
sudo systemctl start segmentationui
-
You can also check for additional applications by listing the installed appnameui.service files installed on the device.
$ ls /usr/lib/systemd/system/*ui.service
/usr/lib/systemd/system/bodyposeui.service /usr/lib/systemd/system/facedetectui.service
/usr/lib/systemd/system/detectionui.service /usr/lib/systemd/system/headposeui.service
/usr/lib/systemd/system/faceblurui.service /usr/lib/systemd/system/segmentationui.service
Comments
0 comments
Please sign in to leave a comment.