# Qualcomm IM SDK release 1.5 Tab Qualcomm Linux Tab Ubuntu - *class* tabincludedirective - ## Release information Table : Software version | Software | Version | | --- | --- | | Yocto | Scarthgap 5.0.9 | | Kernel | 6.6.90 | Table : Release tag version | Release tag | Version | | --- | --- | | Firmware | r1.0\_00095.0 | | Manifest | qcom-6.6.90-QLI.1.5-Ver.1.1 | | Meta-qcom-extras | r1.0\_00097.0 | | Qualcomm IM SDK | qcom-6.6.90-QLI.1.5-Ver.1.1\_qim-product-sdk-2.0.1 | Table : Supported platforms and reference kits | SoC platforms | Reference kits | | --- | --- | | QCS6490 | | | QCS5430 | | | Qualcomm Dragonwing^TM^ IQ-9075 | | | Qualcomm Dragonwing^TM^ IQ-8275 | Qualcomm Dragonwing^TM^ IQ-8275 Beta Evaluation Kit (EVK) | ## Contents of the release The contents of the Qualcomm IM SDK release include: - Recipes for building the individual components: - Qualcomm^®^ Intelligent Multimedia SDK (IM SDK) - Lite Runtime (LiteRT) - Qualcomm^®^ Neural Processing SDK - Qualcomm^®^ AI Engine direct SDK - Sample applications that demonstrate how to use the Qualcomm IM SDK to develop AI edge-based applications. To get started with the Qualcomm IM SDK, see [Qualcomm Intelligent Multimedia (IM SDK) Quick Start Guide](https://docs.qualcomm.com/bundle/publicresource/topics/80-70020-51/qmi-sdk-qsg-landing-page.html). ## New features The following new features applicable to QCS6490, Dragonwing IQ-9075, and Dragonwing IQ-8275 are introduced in the Qualcomm IM SDK release: - AI/ML supports: - People counting in both people-analytics and region-analytics, which provides improved insights into occupancy and movement. - The `qtimlmetaextractor` plugin, which extracts and provides serialized machine learning metadata from the buffer for downstream processing. - AI box supports daisy chain up to 32 streams. - Video supports the real-time messaging protocol (RTMP), which enables real-time video streaming to compatible platforms. ## Sample applications For the complete list of sample applications supported in the Qualcomm IM SDK and instructions on how to run them, see [Sample applications](https://docs.qualcomm.com/bundle/publicresource/topics/80-70020-50/example-applications.html). ## Issues resolved The following issues are resolved in the Qualcomm IM SDK release: - The `gst-ai-parallel-inference` sample application hangs on Dragonwing IQ-8275. - Support for GStreamer daemon (GStd), NV12 universal band width compression (UBWC) compressed streams, and sample applications dependent on NV12 UBWC isn’t enabled inside the docker. ## Limitations The following are the known limitations in the Qualcomm IM SDK release: - When using the `qtioverlay` plugin with detection models, frame drops may occur, especially with many detections. **Workaround**: Use `qtivcomposer` for detection-based ML use cases. - AI/ML parallel inference for 32 streams varies between 20–30 fps depending on the model and input stream, which is less than the expected 30 fps. - Segfault occurs while using Ctrl + C for Qualcomm^®^ Neural Processing use cases with DLC models. - Frame drops are observed with the Qualcomm^®^ Neural Network plugin while running on GPU delegate. - Low FPS is observed with GStreamer pipelines for daisychain and pose detection. - Stability issues are observed with GStreamer pipeline in batched model use cases. - High inference time is observed with `deeplabv3_resnet50.dlc`. - Concurrent streams with smart codec functionality encounter pipeline stalls at the end of the stream. - Reverse playback is limited to video streams with a GOP length that fits within the buffer limitations of the video driver (<= 26). - For Dragonwing IQ-9075 and Dragonwing IQ-8275: > > > - Random corruption is observed at the beginning of playback for low-resolution (480p) NV12 UBWC compressed streams. > - A hang is observed while running use cases with batched models using GStreamer. - For QCS6490: > > > - Performance drops in superresolution end-to-end use cases due to post-processing issues affecting GPU stability. - The `waylandsink` plugin doesn’t support displaying windows at custom screen coordinates. - Starting from Qualcomm Neural Processing SDK version 2.34, LiteRT models encounter failure when executed with the benchmark\_model application using the Qualcomm Neural Network external delegate. - While running the `gst-daisychain-detection-pose` sample application, the output doesn’t display pose information within the extracted ROI regions. - Support for opencv isn’t enabled inside the Qualcomm IM SDK docker container. - The following applications have stability issues due to bugs in the application: - `gst-camera-three-stream-encode-file-detection-display-classification-rtsp.py` - `gst-camera-two-stream-detection-and-classification-side-by-side.py` - Playback of files with dynamic resolution change (DRC) may fail when the final resolution switch is to 4K or 480p. This issue is caused due to the mismatch between the buffer sizes allocated by the application and those expected by the driver. - Use cases involving the `qtismartvencbin` plugin fail due to a missing versioned library. - *class* tabincludedirective - ## Release information Table : Software version | Software | Version | | --- | --- | | Ubuntu | Noble 24.04 | | Kernel | 6.8.0 | Table : Supported platforms and reference kits | SoC platforms | Reference kits | | --- | --- | | QCS6490 | | | QCS5430 | | | Qualcomm Dragonwing^TM^ IQ-9075 | | ## Contents of the release The contents of the Qualcomm IM SDK release include: - Recipes for building the individual components: - Qualcomm IM SDK - Lite Runtime (LiteRT) Note LiteRT was formerly known as TensorFlow Lite. - Qualcomm Neural Processing SDK - Qualcomm AI Engine direct SDK - Sample applications that demonstrate how to use the Qualcomm IM SDK to develop AI edge-based applications. To get started with the Qualcomm IM SDK, see [Qualcomm Intelligent Multimedia (IM SDK) Quick Start Guide](https://docs.qualcomm.com/bundle/publicresource/topics/80-70020-51/qmi-sdk-qsg-landing-page.html). ## New features The following new features are introduced in the Qualcomm IM SDK release: - Microservices: > > > - Yolo V8-based person detection. > - Face detection or recognition for access control. > - General purpose daisy chain. - AI-driven safety and monitoring system. > > > - **Restricted zone alerts**: Define zones in the camera’s FOV to generate alerts when someone enters these areas. > - **PPE compliance**: Ensure safety gear compliance with alerts for missing or restricted items, configurable using a web interface. - The GStreamer version has been upgraded to 1.24. - Qualcomm plugins utilize a new color format, NV12\_Q08C, which is introduced for NV12 with UBWC. This color format replaces the previous method of using flags such as compression=ubwc. - V4L2 supports `dmabuf` and `dmabuf-import` modes for efficient zero-copy data transfer. - Audio AI: End-to-end support for general audio classification including sampling, preprocessing, inference, and post-processing. - The docker base image is updated to Ubuntu 24.04 and the GStreamer version within the docker environment is 1.24. ## Sample applications | Sample application | Description | Supported SoC | | --- | --- | --- | | **AI/ML applications** | **AI/ML applications** | **AI/ML applications** | | `gst-ai-audio-classification` | Classification on streams from audio source. | QCS6490 | | `gst-ai-metadata-parser-example` | Parses metadata using appsink plugin on streams from camera, file, or RTSP sources. Also, it provides the human count from the stream. | QCS6490 | | `gst-ai-usb-camera-app` | USB single camera streaming for preview, video encoder, or network (RTSP) along with object detection and preview. | QCS6490 | | **Video application** | **Video application** | **Video application** | | `gst-opencv-transform` | Video playback using opencv API. Parses input video file, captures frame using cv video capture, and streams videoplay and composites to display using waylandsink. | QCS6490 | | `gst-jpg-image-decode` | Decodes JPEG images where user can view decoded images on waylandsink. | QCS6490 | | `gst-camera-opencv-resize` | Live camera stream using opencv API, which includes color conversion and resize. The application captures camera frame as an input using cv video capture, converts to RGBA using downstream plugin, and resizes to user–specified resolution. Resize output displays on screen. | QCS6490 | | `gst-jpg-decode-example` | Collects the live video input from a camera, file, or an RTSP stream and uses the Qualcomm Neural Network face detection model to produce a preview with the overlaid AI model output on the HDMI display. | QCS6490 | Note `gst-usb-camera-example` is removed from multimedia sample applications. Use `gst-ai-usb-camera-app` instead of `gst-usb-camera-example`. For the complete list of sample applications supported in the Qualcomm IM SDK and instructions on how to run them, see [Sample applications](https://docs.qualcomm.com/bundle/publicresource/topics/80-70020-50/example-applications.html). ## Issues resolved The following issues are resolved in the Qualcomm IM SDK release: - A drop of 1 to 2 fps may be observed with the three-stream camera use case. - Concurrent streams with smart codec functionality encounter cleanup errors at end of stream. - Pose estimation for the `facemap_3dmm_quantized.bin` (Qualcomm Neural Network) AI hub model doesn’t work with the Qualcomm IM SDK pipeline due to caps mismatch. - Caps mismatch with the AI hub face detection Qualcomm Neural Network model. - Support for the FastCV engine is disabled in qtivtransform. ## Limitations The following are the known limitations in the Qualcomm IM SDK release: - Ubuntu desktop doesn’t support sample applications and AI Workflow. - Qualcomm Neural Network use cases don’t work with the `libqnn1` default package. **Workaround**: Upgrade to Qualcomm Neural Network 2.36 (default package is Qualcomm Neural Network 2.35). - The device stops responding when using super resolution use cases on the x04.1 build. - The WebRTC use cases aren’t enabled on this release. - FPS drops to zero while testing multistream AI Camera use cases for long durations. - The following sample applications fail: - `gst-video-playback-example` - `gst-opencv-transform.py` - `gst-jpg-image-decode.py` - `gst-camera-opencv-resize.py` - Camera pipeline hangs while quitting the pipeline of various multistream encode use cases. - The `gst-ai-audio-classification` sample application fails with GPU delegate and displays the error message. > > > ERROR Failed to parse zone configuration > Copy to clipboard ## Related documents - [Qualcomm Intelligent Multimedia SDK (IM SDK) Reference](https://docs.qualcomm.com/bundle/publicresource/topics/80-70020-50/qimsdk_landing_page.html) - [Qualcomm Intelligent Multimedia (IM SDK) Quick Start Guide](https://docs.qualcomm.com/bundle/publicresource/topics/80-70020-51/qmi-sdk-qsg-landing-page.html) - [Qualcomm Dragonwing RB3 Gen 2 Development Kit Guide](https://docs.qualcomm.com/bundle/publicresource/topics/80-70020-251/landing-page.html) - [Qualcomm Dragonwing IQ-9075 Evaluation Kit user guide – Linux](https://docs.qualcomm.com/bundle/publicresource/topics/80-70020-261/iq9-ug-update-the-sw.html) - [Qualcomm Linux Build Guide](https://docs.qualcomm.com/bundle/publicresource/topics/80-70020-254/build_landing_page.html) Last Published: May 14, 2026 [Previous Topic Qualcomm IM SDK release 1.6](https://docs.qualcomm.com/bundle/publicresource/80-80022-52/topics/qim-sdk-release-1-6.md) [Next Topic Qualcomm IM SDK release 1.4](https://docs.qualcomm.com/bundle/publicresource/80-80022-52/topics/qim-sdk-release-1-4.md) Source: [https://docs.qualcomm.com/doc/80-80022-52/topic/qim-sdk-release-1-5.html](https://docs.qualcomm.com/doc/80-80022-52/topic/qim-sdk-release-1-5.html)