# Hardware-accelerated sample applications for AI
Qualcomm Linux provides sample AI/ML applications that show the AI capabilities of the Qualcomm Linux platform.
The following image shows the workflow to run these AI sample apps, from booting the device, downloading the models,
transferring the models to the device to running the sample apps on the target device.
**Workflow for using the sample applications**
The following AI sample applications are part of Qualcomm IM SDK.
For the complete list of supported sample applications, see
[Sample applications](https://docs.qualcomm.com/doc/80-70030-50/topic/example-applications.html)
in the Qualcomm IM SDK documentation.
## AI processing for object detection
The [gst-ai-object-detection](https://docs.qualcomm.com/doc/80-70030-50/topic/gst-ai-object-detection.html)
application allows you to detect objects within images and videos. The use cases show the execution of
[YOLOv5](https://github.com/ultralytics/yolov5), [YOLOv8](https://github.com/ultralytics/ultralytics), and
[YOLOX](https://github.com/Megvii-BaseDetection/YOLOX/blob/main/README.md) using the Qualcomm Neural Processing SDK runtime, YOLOv8 using Qualcomm AI Engine direct, and
YOLOv5 and YOLOv8 using LiteRT.
The following figure shows the pipeline, which receives the input from a live camera feed, file, or an RTSP stream, preprocesses it, runs inferences on
AI hardware, and displays the results on the screen. For information about the plugins used in the pipeline flow, see
[object detection pipeline flow](https://docs.qualcomm.com/doc/80-70030-50/topic/gst-ai-object-detection.html#gst-ai-object-detection__section_p2w_33y_kbc).
## AI processing for multistream-inference
The [gst-ai-multistream-inference](https://docs.qualcomm.com/doc/80-70030-50/topic/multistream-inference.html)
application shows AI inference (object detection and classification) on up to 32 input streams from camera, file, or RTSP stream.
The following figure shows the pipeline, which receives several input streams, preprocesses them, runs AI inferences, combines the
streams, and merges them all into a single video output.
The maximum number of input streams supported on each SoC are follows:
- QCS6490: 8
- Qualcomm Dragonwing IQ-8275: 16
- Qualcomm Dragonwing IQ-9075: 32
Note
IQ-8275 is not supported on Ubuntu Server.
The output is displayed on an HDMI display, saved as an H.264 encoded MP4 file, or converted to an RTSP stream.
For information about the plugins used in this pipeline, see
[multistream-inference pipeline flow](https://docs.qualcomm.com/doc/80-70030-50/topic/multistream-inference.html#multistream-inference__section_opk_tn2_ncc).
## Parallel AI inference
The [gst-ai-parallel-inference](https://docs.qualcomm.com/doc/80-70030-50/topic/gst-ai-parallel-inference.html)
application allows you to perform object detection, object classification, pose detection, and image segmentation on an input stream from different sources,
such as camera, file, or an RTSP network. The use cases implement the LiteRT models for object detection, image segmentation, classification, and pose detection.
Note
This application isn’t supported on IQ-8275.
The following figure shows the pipeline, which receives input streams from a camera, file, or an RTSP stream, performs the parallel inferencing for the four use cases,
and displays the results side by side on the screen.
For information about the plugins used in the pipeline flow, see
[parallel inference pipeline flow](https://docs.qualcomm.com/doc/80-70030-50/topic/gst-ai-parallel-inference.html#gst-ai-parallel-inference__section_gcg_r3s_lbc).
## Daisychain of object detection and pose estimation
The [gst-ai-daisychain-detection-pose](https://docs.qualcomm.com/doc/80-70030-50/topic/daisy-chain-detection-and-pose-detection.html)
application allows you to perform cascaded object detection and pose detection with a camera, file source, or an RTSP stream. The use cases involve detecting
objects and estimating the body poses of the subject in an image or a video.
The following figure show the application workflow, which receives the source, postprocesses it, runs inferences on AI hardware, and displays the results on the screen.
For information about the plugins used in the pipeline flow, see
[daisychain detection and pose pipeline flow](https://docs.qualcomm.com/doc/80-70030-50/topic/daisy-chain-detection-and-pose-detection.html#daisy-chain-detection-and-pose-detection__section_pqq_1ny_kbc).
Last Published: Jun 19, 2026
[Previous Topic
GStreamer ML plugin and AI SDK APIs](https://docs.qualcomm.com/bundle/publicresource/80-70030-15/topics/interfaces.md) [Next Topic
Develop an AI application](https://docs.qualcomm.com/bundle/publicresource/80-70030-15/topics/developer-workflow.md)