Release Notes 2025#

Version 2025.2.0#

Intel® Deep Learning Streamer (Intel® DL Streamer) Pipeline Framework is a streaming media analytics framework, based on GStreamer* multimedia framework, for creating complex media analytics pipelines. It ensures pipeline interoperability and provides optimized media, and inference operations using Intel® Distribution of OpenVINO™ Toolkit Inference Engine backend, across Intel® architecture, CPU, discrete GPU, integrated GPU and NPU. The complete solution leverages:

  • Open source GStreamer* framework for pipeline management

  • GStreamer* plugins for input and output such as media files and real-time streaming from camera or network

  • Video decode and encode plugins, either CPU optimized plugins or GPU-accelerated plugins based on VAAPI

  • Deep Learning models converted from training frameworks TensorFlow*, Caffe* etc.

  • The following elements in the Pipeline Framework repository:

    Element

    Description

    gvaattachroi

    Adds user-defined regions of interest to perform inference on, instead of full frame.

    gvaaudiodetect

    Performs audio event detection using AclNet model.

    gvaaudiotranscribe

    Performs audio transcription using OpenVino GenAI Whisper model.

    gvaclassify

    Performs object classification. Accepts the ROI as an input and outputs classification results with the ROI metadata.

    gvadetect

    Performs object detection on a full-frame or region of interest (ROI) using object detection models such as YOLOv4-v11, MobileNet SSD, Faster-RCNN etc. Outputs the ROI for detected objects.

    gvafpscounter

    Measures frames per second across multiple streams in a single process.

    gvagenai

    Performs inference with Vision Language Models using OpenVINO™ GenAI, accepts video and text prompt as an input, and outputs text description. It can be used to generate text summarization from video.

    gvainference

    Runs deep learning inference on a full-frame or ROI using any model with an RGB or BGR input.

    gvametaaggregate

    Aggregates inference results from multiple pipeline branches

    gvametaconvert

    Converts the metadata structure to the JSON format.

    gvametapublish

    Publishes the JSON metadata to MQTT or Kafka message brokers or files.

    gvamotiondetect

    Performs lightweight motion detection on NV12 video frames and emits motion regions of interest (ROIs) as analytics metadata.

    gvapython

    Provides a callback to execute user-defined Python functions on every frame. Can be used for metadata conversion, inference post-processing, and other tasks.

    gvarealsense

    Provides integration with Intel RealSense cameras, enabling video and depth stream capture for use in GStreamer pipelines.

    gvatrack

    Performs object tracking using zero-term, or imageless tracking algorithms. Assigns unique object IDs to the tracked objects.

    gvawatermark

    Overlays the metadata on the video frame to visualize the inference results.

For the details on supported platforms, please refer to System Requirements. For installing Pipeline Framework with the prebuilt binaries or Docker* or to build the binaries from the open source, refer to Intel® DL Streamer Pipeline Framework installation guide.

New

Title

High-level description

Motion detection (gvamotiondetect)

Performs lightweight motion detection on NV12 video frames and emits motion regions of interest (ROIs) as analytics metadata.

Audio transcription (gvaaudiotranscribe)

Transcribes audio content with OpenVino GenAI Whisper model.

Gvagenai element added

Performs inference with Vision Language Models using OpenVINO™ GenAI, accepts video and text prompt as an input, and outputs text description.
Models supported: MiniCPM-V, Gemma3, Phi-4-multimodal-instruct.

Deep SORT

Preview version of Deep SORT tracking algorithm in gvatrack element.

gvawatermark element support on GPU

Gvawatermark implementation extended about GPU support (CPU default).

Pipeline optimizer support

1st version of DL Streamer optimizer implementation added allowing end user finding the most FPS optimized pipeline.

GstAnalytics metadata support

Enabled GstAnalytics metadata support.

OpenVINO custom operations

Add support for OpenVINO custom operations.

D3D11 preprocessing enabled

Windows support extended about D3D11 preprocessing implementation.

UX, Stability && Performance fixes

• memory management fixes
• automatically select pre-process-backend=va-surface-sharing for GPU
• adjusting caps negotiations and preproc backend selection
• removing deleted element from all shared reference lists.
• using OpenCV preproc to convert sparse tensors to contiguous tensors
• creation of new VADisplay ctx per each inference instance
• remove need for dual va+opencv image pre-processing

Intel Core Ultra Panther Lake CPU/GPU support

Readiness for supporting Intel Core Ultra Panther Lake CPU/GPU.

OpenVINO update

Update to 2025.3 version.

GStreamer update

Update to 1.26.6 version.

GPU drivers update

Update to 25.40 version (for Ubuntu24)

NPU drivers update

Update to 1.23 version.

Fixed

#

Issue Description

1

Fixed issue with segmentation fault and early exit for testing scenarios with mixed GPU/CPU device combinations.

2

Updated documentation for latency tracer.

3

Fixed issue where NPU inference required inefficient CPU color processing.

4

Fixed memory management for elements: gvawatermark, gvametaconvert, gvaclassify.

5

Improved model-proc check logic for va backend.

6

Fixed keypoints metadata processing issue for gvawatermark.

7

Fixed issue with missed gvarealsense element in dlstreamer image.

8

Fixed issue for scenario when vacompositor scale-method option didn’t take affect.

9

Fixed documentation bug in the installation guide.

10

Fixed issue with same name for many python modules used by gvapython.

11

Fixed issue with draw_face_attributes sample (cpp) on TGL Ubuntu 24.

12

Fixed wrong pose estimation on ARL GPU with yolo11s-pose.

13

Fixed inconsistent timestamp for vehicle_pedestrian_tracking sample on ARL.

14

Fixed missing element ‘qsvh264dec’ in Ubuntu24 docker images.

Known Issues

Issue

Issue Description

Preview Architecture 2.0 Samples

Preview Arch 2.0 samples have known issues with inference results.

Sporadic hang on vehicle_pedestrian_tracking_20_cpu sample

Using Tiger Lake CPU to run this sample may lead to sporadic hang at 99.9% of video processing. Rerun the sample as W/A or use GPU instead.

Version 2025.1.2#

Deep Learning Streamer Pipeline Framework is a streaming media analytics framework, based on GStreamer* multimedia framework, for creating complex media analytics pipelines. It ensures pipeline interoperability and provides optimized media, and inference operations using Intel® Distribution of OpenVINO™ Toolkit Inference Engine backend, across Intel® architecture, CPU, discrete GPU, integrated GPU and NPU. The complete solution leverages:

  • Open source GStreamer* framework for pipeline management

  • GStreamer* plugins for input and output such as media files and real-time streaming from camera or network

  • Video decode and encode plugins, either CPU optimized plugins or GPU-accelerated plugins based on VAAPI

  • Deep Learning models converted from training frameworks TensorFlow*, Caffe* etc.

  • The following elements in the Pipeline Framework repository:

    Element

    Description

    gvadetect

    Performs object detection on a full-frame or region of interest (ROI) using object detection models such as YOLOv4-v11, MobileNet SSD, Faster-RCNN etc. Outputs the ROI for detected objects.

    gvaclassify

    Performs object classification. Accepts the ROI as an input and outputs classification results with the ROI metadata.

    gvainference

    Runs deep learning inference on a full-frame or ROI using any model with an RGB or BGR input.

    gvatrack

    Performs object tracking using zero-term, or imageless tracking algorithms. Assigns unique object IDs to the tracked objects.

    gvaaudiodetect

    Performs audio event detection using AclNet model.

    gvagenai

    Performs inference with Vision Language Models using OpenVINO™ GenAI, accepts video and text prompt as an input, and outputs text description. It can be used to generate text summarization from video.

    gvaattachroi

    Adds user-defined regions of interest to perform inference on, instead of full frame.

    gvafpscounter

    Measures frames per second across multiple streams in a single process.

    gvametaaggregate

    Aggregates inference results from multiple pipeline branches

    gvametaconvert

    Converts the metadata structure to the JSON format.

    gvametapublish

    Publishes the JSON metadata to MQTT or Kafka message brokers or files.

    gvapython

    Provides a callback to execute user-defined Python functions on every frame. Can be used for metadata conversion, inference post-processing, and other tasks.

    gvarealsense

    Provides integration with Intel RealSense cameras, enabling video and depth stream capture for use in GStreamer pipelines.

    gvawatermark

    Overlays the metadata on the video frame to visualize the inference results.

For the details on supported platforms, please refer to System Requirements. For installing Pipeline Framework with the prebuilt binaries or Docker* or to build the binaries from the open source, refer to Deep Learning Streamer Pipeline Framework installation guide.

New

Title

High-level description

Custom model post-processing

End user can now create a custom post-processing library (.so); sample added as reference.

Latency mode support

Default scheduling policy for DL Streamer is throughput. With this change user can add scheduling-policy=latency for scenarios that prioritize latency requirements over throughput.

Visual Embeddings enabled

New models enabled to convert input video into feature embeddings, validated with Clip-ViT-Base-B16/Clip-ViT-Base-B32 models; sample added as reference.

VLM models support

new gstgenai element added to convert video into text (with VLM models), validated with miniCPM2.6, available in advanced installation option when building from sources; sample added as reference.

INT8 automatic quantization support for Yolo models

Performance improvement, automatic INT8 quantization for Yolo models

MS Windows 11 support

Native support for Windows 11

New Linux distribution (Azure Linux derivative)

New distribution added, DL Streamer can be now installed on Edge Microvisor Toolkit.

License plate recognition use case support

Added support for models that allow to recognize license plates; sample added as reference.

Deep Scenario model support

Commercial 3D model support

Anomaly model support

Added support for anomaly model, sample added as reference, sample added as reference.

RealSense element support

New gvarealsense element implementation providing basic integration with Intel RealSense cameras, enabling video and depth stream capture for use in GStreamer pipelines.

OpenVINO 2025.3 version support

Support of recent OpenVINO version added.

GStreamer 1.26.6 version support

Support of recent GStreamer version added.

NPU 1.19 version driver support

Support of recent NPU driver version added.

Docker image size reduction

Reduction for all images, e.g., Ubuntu 24 Release image size reduced to 1.6GB from 2.6GB

Known Issues

Issue

Issue Description

VAAPI memory with decodebin

If you are using decodebin in conjunction with vaapi-surface-sharing preprocessing backend you should set caps filter using ""video/x-raw(memory:VASurface)"" after decodebin to avoid issues with pipeline initialization

Artifacts on sycl_meta_overlay

Running inference results visualization on GPU via sycl_meta_overlay may produce some partially drawn bounding boxes and labels

Preview Architecture 2.0 Samples

Preview Arch 2.0 samples have known issues with inference results.

Sporadic hang on vehicle_pedestrian_tracking_20_cpu sample

Using Tiger Lake CPU to run this sample may lead to sporadic hang at 99.9% of video processing. Rerun the sample as W/A or use GPU instead.

Simplified installation process for option 2 via script

In certain configurations, users may encounter visible errors

Error when using legacy YoloV5 models: Dynamic resize: Model width dimension shall be static

To avoid the issue, modify samples/download_public_models.sh by inserting the following snippet at lines 273 and 280:

python3 - <<EOF “”${MODEL_NAME}””
import sys, os
from openvino.runtime import Core
from openvino.runtime import save_model
model_name = sys.argv[1]
core = Core()
os.rename(f””{model_name}_openvino_model””, f””{model_name}_openvino_modelD””)
model = core.read_model(f””{model_name}_openvino_modelD/{model_name}.xml””)
model.reshape([-1, 3, 640, 640])

Version 2025.0.1.3#

Intel® Deep Learning Streamer (Intel® DL Streamer) Pipeline Framework is a streaming media analytics framework, based on GStreamer* multimedia framework, for creating complex media analytics pipelines. It ensures pipeline interoperability and provides optimized media, and inference operations using Intel® Distribution of OpenVINO™ Toolkit Inference Engine backend, across Intel® architecture, CPU, discrete GPU, integrated GPU and NPU.

This release includes DL Streamer Pipeline Framework elements to enable video and audio analytics capabilities, (e.g., object detection, classification, audio event detection), and other elements to build end-to-end optimized pipeline in GStreamer* framework.

The complete solution leverages:

  • Open source GStreamer* framework for pipeline management

  • GStreamer* plugins for input and output such as media files and real-time streaming from camera or network

  • Video decode and encode plugins, either CPU optimized plugins or GPU-accelerated plugins based on VAAPI

  • Deep Learning models converted from training frameworks TensorFlow*, Caffe* etc.

  • The following elements in the Pipeline Framework repository:

Element

Description

gvadetect

Performs object detection on a full-frame or region of interest (ROI) using object detection models such as YOLOv4-v11, MobileNet SSD, Faster-RCNN etc. Outputs the ROI for detected objects.

gvaclassify

Performs object classification. Accepts the ROI as an input and outputs classification results with the ROI metadata.

gvainference

Runs deep learning inference on a full-frame or ROI using any model with an RGB or BGR input.

gvatrack

Performs object tracking using zero-term, or imageless tracking algorithms. Assigns unique object IDs to the tracked objects.

gvaaudiodetect

Performs audio event detection using AclNet model.

gvaattachroi

Adds user-defined regions of interest to perform inference on, instead of full frame.

gvafpscounter

Measures frames per second across multiple streams in a single process.

gvametaaggregate

Aggregates inference results from multiple pipeline branches

gvametaconvert

Converts the metadata structure to the JSON format.

gvametapublish

Publishes the JSON metadata to MQTT or Kafka message brokers or files.

gvapython

Provides a callback to execute user-defined Python functions on every frame. Can be used for metadata conversion, inference post-processing, and other tasks.

gvawatermark

Overlays the metadata on the video frame to visualize the inference results.

For the details of supported platforms, please refer to System Requirements section.

For installing Pipeline Framework with the prebuilt binaries or Docker* or to build the binaries from the open source, please refer to DL Streamer Pipeline Framework installation guide

New

Title

High-level description

Installation process

Enhanced installation scripts for the ‘installation on host’ option

Post installation steps

Added a selection option for the YOLO model and device to the hello_dlstreamer.sh script

Download models

Improved download_public_models.sh script

Documentation updates

Improved installation processes descriptions and tutorial refresh

Known Issues

Issue

Issue Description

VAAPI memory with decodebin

If you are using decodebin in conjunction with vaapi-surface-sharing preprocessing backend you should set caps filter using "video/x-raw(memory:VASurface)" after decodebin to avoid issues with pipeline initialization

Artifacts on sycl_meta_overlay

Running inference results visualization on GPU via sycl_meta_overlay may produce some partially drawn bounding boxes and labels

Preview Architecture 2.0 Samples

Preview Arch 2.0 samples have known issues with inference results

Sporadic hang on vehicle_pedestrian_tracking_20_cpu sample

Using Tiger Lake CPU to run this sample may lead to sporadic hang at 99.9% of video processing, rerun the sample as W/A or use GPU instead

Simplified installation process for option 2 via script

In certain configurations, users may encounter visible errors

Error when using legacy YoloV5 models: Dynamic resize: Model width dimension shall be static

To avoid the issue, modify samples/download_public_models.sh by inserting the following snippet at lines 273 and 280:

python3 - <<EOF “${MODEL_NAME}”
import sys, os
from openvino.runtime import Core
from openvino.runtime import save_model
model_name = sys.argv[1]
core = Core()
os.rename(f”{model_name}_openvino_model”, f”{model_name}_openvino_modelD”)
model = core.read_model(f”{model_name}_openvino_modelD/{model_name}.xml”)
model.reshape([-1, 3, 640, 640])

Version 2025.0.2#

New

Title

High-level description

Geti Models 2.7 version

Support for Geti Classification/Detection Models in 2.7 version

GStreamer plugins

Support for gst-rswebrtc-plugins

Documentation updates

Documentation updates - “queue” element

Version 2025.0.1#

New

Title

High-level description

LVM support

Support for Large Vision Models

LVM support

Sample demonstrating image embedding extraction with Visual Transformer (LVM)

OpenVINO 2025.0 support

Update to the latest version of OpenVINO

GStreamer 1.24.12 support

Update GStreamer to 1.24.12 version

Updated NPU driver

Updated NPU driver to 1.13.0 version.

Documentation updates

Documentation how to convert from DeepStream to Deep Learning Steamer

Version 2025.0.0#

New

Title

High-level description

Enhanced support of Intel® Core™ Ultra Processors (Series 2) (formerly codenamed Lunar Lake); enabled va-surface-sharing pre-process backend.

Validated with Ubuntu 24.04, 6.12.3-061203-generic and the latest Intel® Graphics Compute Runtime for oneAPI Level Zero and OpenCL™ Driver v24.52.32224.5

[preview] Enabled Intel® Arc™ B-Series Graphics [products formerly Battlemage]

Validated with Ubuntu 24.04, 6.12.3-061203-generic and the latest Intel® Graphics Compute Runtime for oneAPI Level Zero and OpenCL™ Driver v24.52.32224.5 + the latest public Intel Graphics Media Driver version + pre-rerelease Intel® Graphics Memory Management Library version

OpenVINO 2024.6 support

Update to the latest version of OpenVINO

Updated NPU driver

Updated NPU driver to 1.10.1 version.

Bug fixing

Running multiple gstreamer pipeline objects in the same process on dGPU leads to error; DL Streamer docker image build is failing (2024.2.2 and 2024.3.0 versions); Fixed installation scripts: minor fixes of GPU, NPU installation section; Updated documentation: cleanup, added missed parts, added DLS system requirements