# Sample applications
Qualcomm Linux includes various sample applications. Instructions for running four of these sample applications have been provided for your quick reference. For more information, see [Sample
applications](https://docs.qualcomm.com/bundle/publicresource/topics/80-70017-50/example-applications.html).
There are two main groups of sample applications available, each suited for different use cases such as retail, Edge AI box, and IP camera.
[Multimedia applications](https://docs.qualcomm.com/doc/80-70017-253/topic/demo_app.html#link-mm-apps)
These are related to camera, video, and audio functionalities.
[AI applications](https://docs.qualcomm.com/doc/80-70017-253/topic/demo_app.html#link-ai-apps)
These focus on AI and machine learning (ML) capabilities.
Note
- To run the multimedia and AI applications, ensure that the [Wi-Fi](https://docs.qualcomm.com/doc/80-70017-253/topic/set_up_the_device.html#using-wifi) is set up and [SSH](https://docs.qualcomm.com/doc/80-70017-253/topic/ubuntu_host.html#ssh-connection) connectivity is established.
- To view the display output, connect the HDMI display to the HDMI port of the RB3 Gen 2 device (see [Connect to HDMI display](https://docs.qualcomm.com/doc/80-70017-253/topic/set_up_the_device.html#concept-wc5-hcp-4bc)).
- If you want to run sample applications from the UART shell, remount the file system with read/write permission using the following command:
mount -o remount, rw /usr
Copy to clipboard
## Multimedia applications
The multimedia sample applications demonstrate use cases for camera, display, and video streams on the RB3 Gen 2 device.
### Multicamera streaming or encoding (Dash cam)
The **gst-multi-camera-example** command-line application demonstrates simultaneous streaming from two camera sensors on the RB3 Gen 2 device. The application composes the camera feeds side-by-side to display on a screen or encodes and stores the video streams to files.
**Figure: Dash cam application workflow**
**Video: Multicamera application**
**Example usage**
To execute the application, run the following use cases in the SSH shell:
1. To view the sample application on the HDMI display, run the following export command:
export XDG_RUNTIME_DIR=/dev/socket/weston && export WAYLAND_DISPLAY=wayland-1
Copy to clipboard
2. To view the `waylandsink` output, run the following command:
gst-multi-camera-example -o 0
Copy to clipboard
3. To store the encoder output, do the following:
1. Run the following command:
gst-multi-camera-example -o 1
Copy to clipboard
The encoded files are stored at `/opt/cam1_vid.mp4` and
`/opt/cam2_vid.mp4` for camera 1 and camera 2, respectively.
2. To pull the files from the host machine, run the following command:
scp root@:/opt/cam1_vid.mp4
Copy to clipboard
Note
When prompted for a password, enter `oelinux123`.
3. To play the encoder output, you can use any media player that supports MP4 files.
- To stop the use case, press **CTRL + C**.
- To display the available help options, run the following command:
gst-multi-camera-example --help
Copy to clipboard
- The GStreamer debug output is controlled by the `GST_DEBUG` environment variable. Set the required level to enable logging. For example, to log all warnings, run the following command:
export GST_DEBUG=2
Copy to clipboard
### Multichannel video decode and compose (Video wall)
The **gst-concurrent-videoplay-composition** command-line application facilitates concurrent video decode and playback for advanced video
coding (AVC)-coded videos and performs composition on a display for video wall application. The application requires at least one input
video file, which is expected to be an MP4 file with the AVC codec.
**Figure: Video wall application workflow**
**Video: Multichannel decode and display application**
**Example usage**
1. To transfer prerecorded or test videos that are in the AVC-encoded MP4 (H.264) format (with the filename as ``) to your device, run the following command on the host:
scp root@[DEVICE IP-ADDR]:/opt/
Copy to clipboard
Note
When prompted for a password, enter `oelinux123`.
2. To view the sample application on the HDMI display, run the following export command in the SSH shell:
export XDG_RUNTIME_DIR=/dev/socket/weston && export WAYLAND_DISPLAY=wayland-1
Copy to clipboard
3. To experience concurrent playback for four channels, run the following command:
gst-concurrent-videoplay-composition -c 4 -i /opt/.mp4 -i /opt/.mp4 -i /opt/.mp4 -i /opt/.mp4
Copy to clipboard
Note
- `-c`: specifies that the number of streams to be decoded for composition can be either 2, 4, or 8.
- `-i`: specifies the absolute path to the input video file.
- To stop the use case, press **CTRL + C**.
- To display the available help options, run the following command:
gst-concurrent-videoplay-composition --help
Copy to clipboard
- The GStreamer debug output is controlled by the `GST_DEBUG` environment variable. Set the required level to enable logging. For example, to log all warnings, run the following command:
export GST_DEBUG=2
Copy to clipboard
## AI applications
AI sample applications demonstrate use cases for object detection and parallel inferencing on input streams from a camera, video file, or RTSP stream on the RB3 Gen 2 device. To experience these sample applications, you must obtain AI models from [Qualcomm AI Hub](https://aihub.qualcomm.com/iot/models) and labels from GitHub.
The procedure involves downloading the models and labels, transferring them to the RB3 Gen 2 device, and running the sample applications.
**Figure: Workflow for running AI sample applications**
**Prerequisite**
AI sample applications require model and label files on the device to run the application.
**Procedure**
1. Download the models from [Qualcomm AI Hub](https://aihub.qualcomm.com/iot/models).
| Sample application | Models required |
| --- | --- |
| Object detection | [Yolov8-Detection-Quantized](https://github.com/quic/ai-hub-models/tree/main/qai_hub_models/models/yolov8_det_quantized) |
| Parallel AI inference | [Yolov8-Detection-Quantized](https://github.com/quic/ai-hub-models/tree/main/qai_hub_models/models/yolov8_det_quantized) |
| Parallel AI inference | [Inception-v3-Quantized](https://aihub.qualcomm.com/iot/models/inception_v3_quantized) |
| Parallel AI inference | [HRNetPoseQuantized](https://aihub.qualcomm.com/iot/models/hrnet_pose_quantized) |
| Parallel AI inference | [DeepLabV3-Plus-MobileNet-Quantized](https://aihub.qualcomm.com/iot/models/deeplabv3_plus_mobilenet_quantized) |
| | |
| | |
| | |
Note
The YOLOv8 model is not available by default. You need to export the model using the AI Hub APIs. For instructions, see [YOLOv8-Detection-Quantized](https://github.com/quic/ai-hub-models/tree/main/qai_hub_models/models/yolov8_det_quantized).
2. Update the `q_offset` and `q_scale` constants of the quantized LiteRT model in the JSON file. For instructions, see [Obtain model constants](https://docs.qualcomm.com/bundle/publicresource/topics/80-70017-15B/integrate-ai-hub-models.html#obtain-model-constants).
3. Use the following command to push the downloaded model files to the device:
scp root@:/opt/
Copy to clipboard
**Example**
scp mobilenet_v2_quantized.tflite root@:/opt/
Copy to clipboard
4. On the target device, obtain the download_artifacts.sh script, set executable permissions, and run it with the required arguments to download the model and label files to the device.
curl -L -O https://raw.githubusercontent.com/quic/sample-apps-for-qualcomm-linux/refs/heads/main/qualcomm-linux/scripts/download_artifacts.sh
Copy to clipboard
chmod +x download_artifacts.sh
Copy to clipboard
./download_artifacts.sh
Copy to clipboard
### AI object detection
The **gst-ai-object-detection** sample application demonstrates the hardware capability to perform object detection on input streams from a camera, video file, or RTSP stream. The pipeline receives the input stream, performs preprocessing, runs inferences on AI hardware, and displays the results on the screen.
**Figure: Object detection application workflow**
**Video: Object detection application**
**Example usage**
It is mandatory to push the model and label files to the device to run the sample application. For details, see [Prerequisite](https://docs.qualcomm.com/doc/80-70017-253/topic/demo_app.html#prereq-ai).
1. To view the sample application on the HDMI display, run the following export command in the SSH shell:
export XDG_RUNTIME_DIR=/dev/socket/weston && export WAYLAND_DISPLAY=wayland-1
Copy to clipboard
2. Edit the `/opt/config_detection.json` file on your device.
{
"camera": 0,
"ml-framework": "tflite",
"yolo-model-type": "yolov8",
"model": "/opt/YOLOv8-Detection-Quantized.tflite",
"labels": "/opt/yolonas.labels",
"constants": "YOLOv8,q-offsets=<21.0, 0.0, 0.0>,q-scales=<3.093529462814331, 0.00390625, 1.0>;",
"threshold": 40,
"runtime": "dsp"
}
Copy to clipboard
3. To execute the application, run the following command:
gst-ai-object-detection
Copy to clipboard
- To stop the use case, press **CTRL + C**.
- To display the available help options, run the following command:
gst-ai-object-detection -h
Copy to clipboard
- The GStreamer debug output is controlled by the `GST_DEBUG` environment variable. Set the required level to enable logging. For example, to log all warnings, run the following command:
export GST_DEBUG=2
Copy to clipboard
### Parallel AI fusion
The **gst-ai-parallel-inference** command-line application demonstrates the hardware capability to perform four parallel AI inferences on input streams from a camera, video file, or RTSP stream. The pipeline performs object detection, object classification, pose detection, and image segmentation on the input stream. The results are displayed side-by-side on the screen.
**Figure: Parallel inference application workflow**
**Video: Parallel inferencing application**
**Example usage**
It is mandatory to push the model and label files to the device to run the sample application. For details, see [Prerequisite](https://docs.qualcomm.com/doc/80-70017-253/topic/demo_app.html#prereq-ai).
1. To view the sample application on the HDMI display, run the following export command in the SSH shell:
export XDG_RUNTIME_DIR=/dev/socket/weston && export WAYLAND_DISPLAY=wayland-1
Copy to clipboard
2. To execute the application, run the following command:
gst-ai-parallel-inference
Copy to clipboard
- To stop the use case, press **CTRL + C**.
- To display the available help options, run the following command:
gst-ai-parallel-inference -h
Copy to clipboard
AI Hub frequently updates models with the latest SDK versions. Using incorrect model constants might lead to inaccurate results. If you encounter such issues, update the model constants. You can provide the model constants for the sample application using the following command:
gst-ai-parallel-inference \
--object-detection-constants="YOLOv8,q-offsets=<21.0, 0.0, 0.0>,q-scales=<3.093529462814331, 0.00390625, 1.0>;" \
--pose-detection-constants="Posenet,q-offsets=<8.0>,q-scales=<0.0040499246679246426>;" \
--segmentation-constants="deeplab,q-offsets=<61.0>,q-scales=<0.06232302635908127>;" \
--classification-constants="Mobilenet,q-offsets=<-95.0>,q-scales=<0.18740029633045197>;"
Copy to clipboard
- The GStreamer debug output is controlled by the `GST_DEBUG` environment variable. Set the required level to enable logging. For example, to log all warnings, run the following command:
export GST_DEBUG=2
Copy to clipboard
**Known issue**
In pose detection, the model detects only one person, even if multiple people are present in the frame.
Note
Image classification using the Inception v3 model is trained on the ImageNet data set. As a result, the model cannot detect a person because this class is not included in the data set.
## More apps
The Qualcomm Linux release includes a wider variety of sample applications. To explore and experience additional applications, see [Sample applications](https://docs.qualcomm.com/bundle/publicresource/topics/80-70017-50/example-applications.html).
## Develop applications for Qualcomm Linux
You can develop applications using the Qualcomm Intelligent Multimedia Product (QIMP) SDK. To get started with your first application, see
[Develop applications](https://docs.qualcomm.com/bundle/publicresource/topics/80-70017-51/application-development.html).
Last Published: Jan 09, 2026
[Previous Topic
Host machine options](https://docs.qualcomm.com/bundle/publicresource/80-70017-253/topics/set_up_the_device.md) [Next Topic
Additional setup](https://docs.qualcomm.com/bundle/publicresource/80-70017-253/topics/additional_setup.md)
Source: [https://docs.qualcomm.com/doc/80-70017-253/topic/demo_app.html](https://docs.qualcomm.com/doc/80-70017-253/topic/demo_app.html)