Edit

Share via

Download Azure Kinect Body Tracking SDK

  • Article

This document provides links to install each version of the Azure Kinect Body Tracking SDK.

Azure Kinect Body Tracking SDK contents

  • Headers and libraries to build a body tracking application using the Azure Kinect DK.
  • Redistributable DLLs needed by body tracking applications using the Azure Kinect DK.
  • Sample body tracking applications.
Version Download
1.1.2 msi nuget
1.1.1 msi nuget
1.1.0 msi
1.0.1 msi nuget
1.0.0 msi nuget

Linux installation instructions

Currently, the only supported distribution is Ubuntu 18.04 and 20.04. To request support for other distributions, see this page.

First, you'll need to configure Microsoft's Package Repository, following the instructions here.

The libk4abt<major>.<minor>-dev package contains the headers and CMake files to build against libk4abt. The libk4abt<major>.<minor> package contains the shared objects needed to run executables that depend on libk4abt as well as the example viewer.

The basic tutorials require the libk4abt<major>.<minor>-dev package. To install it, run

sudo apt install libk4abt<major>.<minor>-dev

If the command succeeds, the SDK is ready for use.

Note

When installing the SDK, remember the path you install to. For example, "C:\Program Files\Azure Kinect Body Tracking SDK 1.0.0". You will find the samples referenced in articles in this path. Body tracking samples are located in the body-tracking-samples folder in the Azure-Kinect-Samples repository. You will find the samples referenced in articles here.

Change log

v1.1.2

  • [Feature] Added C# wrapper support for Linux Link
  • [Bug Fix] k4abt_simple_3d_viewer.exe works with latest NVIDIA drivers Link

v1.1.1

  • [Feature] Added cmake support to all body tracking samples
  • [Feature] NuGet package returns. Developed new NuGet package that includes Microsoft developed body tracking dlls and headers, and ONNX runtime dependencies. The package no longer includes the NVIDIA CUDA and TRT dependencies. These continue to be included in the MSI package.
  • [Feature] Upgraded to ONNX Runtime v1.10. Recommended NVIDIA driver version is 472.12 (Game Ready) or 472.84 (Studio). There are OpenGL issues with later drivers.
  • [Bug Fix] CMake missing from offline_processor sample Link
  • [Bug Fix] CPU mode no longer requires NVIDIA CUDA dependencies Link
  • [Bug Fix] Verified samples compile with Visual Studio 2022 and updated samples to use this release Link
  • [Bug Fix] Added const qualifier to APIs Link
  • [Bug Fix] Added check for nullptr handle in shutdown() Link
  • [Bug Fix] Improved dependencies checks Link
  • [Bug Fix] Updated REDIST.TXT file Link
  • [Bug Fix] Improved DirectML performance Link
  • [Bug Fix] Fixed exception declaration in frame::get_body() Link
  • [Bug Fix] Fixed memory leak Link
  • [Bug Fix] Updated dependencies list Link

v1.1.0

  • [Feature] Add support for DirectML (Windows only) and TensorRT execution of pose estimation model. See FAQ on new execution environments.
  • [Feature] Add model_path to k4abt_tracker_configuration_t struct. Allows users to specify the pathname for pose estimation model. Defaults to dnn_model_2_0_op11.onnx standard pose estimation model located in the current directory.
  • [Feature] Include dnn_model_2_0_lite_op11.onnx lite pose estimation model. This model trades ~2x performance increase for ~5% accuracy decrease.
  • [Feature] Verified samples compile with Visual Studio 2019 and updates samples to use this release.
  • [Breaking Change] Update to ONNX Runtime 1.6 with support for CPU, CUDA 11.1, DirectML (Windows only), and TensorRT 7.2.1 execution environments. Requires NVIDIA driver update to R455 or better.
  • [Breaking Change] No NuGet install.
  • [Bug Fix] Add support for NVIDIA RTX 30xx series GPUs Link
  • [Bug Fix] Add support for AMD and Intel integrated GPUs (Windows only) Link
  • [Bug Fix] Update to CUDA 11.1 Link
  • [Bug Fix] Update to Sensor SDK 1.4.1 Link

v1.0.1

  • [Bug Fix] Fix issue that the SDK crashes if loading onnxruntime.dll from path on Windows build 19025 or later: Link

v1.0.0

  • [Feature] Add C# wrapper to the msi installer.
  • [Bug Fix] Fix issue that the head rotation cannot be detected correctly: Link
  • [Bug Fix] Fix issue that the CPU usage goes up to 100% on Linux machine: Link
  • [Samples] Add two samples to the sample repo. Sample 1 demonstrates how to transform body tracking results from the depth space to color space Link; sample 2 demonstrates how to detect floor plane Link

Next steps