Skip to content

Changelog

v1.8 (Jun 28, 2025)

  • Renamed EImageRotation to EThextureProcessImageRotation for UE5.6 and later only. This change was made in response to the addition of EImageRotation of the same name to the VirtualProduction engine plugin.

v1.7 (Jun 23, 2023)

  • Unreal Engine 5.2 version is now available.
  • Added DLLEXPORT so that FOnnxTensorInfo can be used from C++ in other modules.
  • Added conditional compilation to work with the latest API of ONNX Runtime. See Use latest ONNX Runtime for more details.

v1.6 (Jan 27, 2023)

  • Added support for Unreal Engine 5.1.
  • Added OnnxModel destructor call when UOnnxModelWrapper is destroyed.

v1.5 (Apr 10, 2022)

v1.4 (Mar 18, 2022)

v1.3 (Mar 04, 2022)

v1.2 (Feb 18, 2022)

  • Updated TextureProcessing module
    • Added a component to convert UTexture to float array. (TextureProcessFloatComponent)
    • Added functions to create UTexture from arrays of byte or float.
    • Fixed a bug that some UTexture cannot be processed by TextureProcessComponent.
      • Now BP_TextureProcessComponent is deprecated. Use TextureProcessComponent instead.
  • Updated CustomizedOpenCV module
    • Removed OpenCV's check function to avoid conflict with UE4's check macro.
  • Added example projects

v1.1 (Feb 11, 2022)

  • Added support for Ubuntu 18.04.6 Desktop 64bit
    • GPU accelerations by CUDA and TensorRT supported.
      • You need an NVIDIA GPU which supports CUDA, cuDNN, and TensorRT.
      • You need to install CUDA ver 11.4.2, cuDNN ver 8.2.4, and TensorRT ver 8.2.3.0.
      • DNN models which contain unsupported operators cannot be loaded when TensorRT is enabled.
        See the official document for supported operators. (NNEngine uses TensorRT 8.2 as backend on Linux)
    • Tested environment:
      • Unreal Engine: 4.26.2, 4.27.2
      • Vulkan utils: 1.1.70+dfsg1-1ubuntu0.18.04.1
      • .NET SDK: 6.0.101-1
      • OS: Ubuntu 18.04.6 Desktop 64bit
      • CPU: Intel i3-8350K
      • GPU: NVIDIA GeForce GTX 1080 Ti
        • Driver: 470.130.01
        • CUDA: 11.4.2-1
        • cuDNN: 8.2.4
        • TensorRT: 8.2.3.0
  • Added EXPERIMENTAL support for Android as target build
    • Tested environment:
      • Device: Xiaomi Redmi Note 9S
      • Android version: 10 QKQ1.191215.002
    • Note:
      • You need to convert your model to ORT format.
        See the official document for the details.
      • There are some DNN models which cannot be loaded on Android.
      • NNEngine uses ONNX Runtime Mobile ver 1.8.1 on Android.
    • GPU acceleration by NNAPI is not tested yet.

v1.0 (Dec 21, 2021)

  • First release.