diff options
Diffstat (limited to 'docs')
-rw-r--r-- | docs/00_introduction.dox | 31 |
1 files changed, 30 insertions, 1 deletions
diff --git a/docs/00_introduction.dox b/docs/00_introduction.dox index 67b879c37b..55c5fcb2b8 100644 --- a/docs/00_introduction.dox +++ b/docs/00_introduction.dox @@ -253,8 +253,37 @@ v20.05 Public major release - @ref NEDepthConvertLayerKernel - @ref NEDepthConvertLayer - @ref NEGEMMConvolutionLayer - - @ref NEGEMMConvolutionLayer - @ref NEGEMMAssemblyDispatch + - Added new data type QASYMM8_SIGNED support for: + - @ref CLDirectConvolutionLayer + - @ref CLDeconvolutionLayer + - @ref CLDirectDeconvolutionLayer + - @ref CLGEMMDeconvolutionLayer + - @ref CLGEMMLowpMatrixMultiplyReshapedKernel + - @ref CLGEMMLowpQuantizeDownInt32ScaleKernel + - @ref CLGEMMLowpQuantizeDownInt32ScaleByFloatKernel + - @ref CLReductionOperation + - @ref CLReduceMean + - @ref NEUpsampleLayer + - @ref NECast + - @ref NEReductionOperation + - @ref NEReduceMean + - @ref NEArgMinMaxLayer + - @ref NEDeconvolutionLayer + - @ref NEGEMMLowpQuantizeDownInt32ScaleKernel + - @ref CPPBoxWithNonMaximaSuppressionLimit + - @ref CPPDetectionPostProcessLayer + - @ref CPPPermuteKernel + - @ref CPPPermute + - @ref CPPTopKVKernel + - @ref CPPTopKV + - Deprecated OpenCL kernels / functions: + - CLGEMMLowpQuantizeDownInt32ToUint8Scale + - CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFloat + - Deprecated NEON kernels / functions: + - NEGEMMLowpQuantizeDownInt32ToUint8Scale + - Removed CPP kernels / functions: + - CPPFlipWeightsKernel v20.02.1 Maintenance release - Added Android-NN build script. |