Incorrect output after GPU to GPU inference via VideoFrame and Gray8 …

…models (#10425) * If the tensor is of gray8 format, we should call the gray8 shader * other check (which resolves to unknown in this case) is incorrectly being compared to constant and not DXGI_FORMAT Co-authored-by: Sheil Kumar <[email protected]>
microsoft · Jan 28, 2022 · 2dd5e75 · 2dd5e75
1 parent feae842
commit 2dd5e75
Showing 1 changed file with 2 additions and 1 deletion.
diff --git a/winml/lib/Api.Image/TensorToVideoFrameConverter.cpp b/winml/lib/Api.Image/TensorToVideoFrameConverter.cpp
@@ -483,7 +483,8 @@ void TensorToVideoFrameConverter::ConvertGPUTensorToDX12Texture(
   PipelineStateCacheFormat formatFrom = PipelineStateCacheFormat::kBGR8;
   if (tensorDesc.channelType == kImageTensorChannelTypeRGB8) {
     formatFrom = PipelineStateCacheFormat::kRGB8;
-  } else if (inputDesc.Format == kImageTensorChannelTypeGRAY8) {
+  } else if (inputDesc.Format == DXGI_FORMAT_R8_UNORM ||
+             tensorDesc.channelType == kImageTensorChannelTypeGRAY8) {
     formatFrom = PipelineStateCacheFormat::kGRAY8;
   }