/// /// Copyright (c) 2017-2021 Arm Limited. /// /// SPDX-License-Identifier: MIT /// /// Permission is hereby granted, free of charge, to any person obtaining a copy /// of this software and associated documentation files (the "Software"), to /// deal in the Software without restriction, including without limitation the /// rights to use, copy, modify, merge, publish, distribute, sublicense, and/or /// sell copies of the Software, and to permit persons to whom the Software is /// furnished to do so, subject to the following conditions: /// /// The above copyright notice and this permission notice shall be included in all /// copies or substantial portions of the Software. /// /// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR /// IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, /// FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE /// AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER /// LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, /// OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE /// SOFTWARE. /// namespace arm_compute { /** @page versions_changelogs Release Versions and Changelog @tableofcontents @section S2_1_versions Release versions All releases are numbered vYY.MM Where YY are the last two digits of the year, and MM the month number. If there is more than one release in a month then an extra sequential number is appended at the end: v17.03 (First release of March 2017) v17.03.1 (Second release of March 2017) v17.04 (First release of April 2017) @note We're aiming at releasing one major public release with new features per quarter. All releases in between will only contain bug fixes. @section S2_2_changelog Changelog v21.05 Public major release - Removed computer vision support from Arm® Neon™ backend - Removed the following functions: - NEAbsoluteDifference - NEAccumulate - NEBox3x3 - NECannyEdge - NEChannelCombine - NEChannelExtract - NEColorConvert - NEConvolution - NEDerivative - NEDilate - NEEqualizeHistogram - NEErode - NEFastCorners - NEGaussian3x3 - NEGaussian5x5 - NEGaussianPyramid - NEHOGDescriptor - NEHOGDetector - NEHOGGradient - NEHOGMultiDetection - NEHarrisCorners - NEHistogram - NEIntegralImage - NELaplacianPyramid - NELaplacianReconstruct - NEMagnitude - NEMeanStdDev - NEMedian3x3 - NEMinMaxLocation - NENonLinearFilter - NEOpticalFlow - NEPhase - NEScharr3x3 - NESobel3x3 - NESobel5x5 - NESobel7x7 - NETableLookup - NEThreshold - NEWarpAffine - NEWarpPerspectiveKernel - Remove all GLES kernels / functions / tests / examples - Removed computer vision support from CL backend - Removed the following functions: - CLAbsoluteDifference - CLAccumulate - CLBox3x3 - CLCannyEdge - CLChannelCombine - CLChannelExtract - CLColorConvert - CLConvolution - CLDerivative - CLDilate - CLEqualizeHistogram - CLErode - CLFastCorners - CLGaussian3x3 - CLGaussian5x5 - CLGaussianPyramid - CLHOGDescriptor - CLHOGDetector - CLHOGGradient - CLHOGMultiDetection - CLHarrisCorners - CLHistogram - CLIntegralImage - CLLaplacianPyramid - CLLaplacianReconstruct - CLMagnitude - CLMeanStdDev - CLMedian3x3 - CLMinMaxLocation - CLNonLinearFilter - CLOpticalFlow - CLPhase - CLScharr3x3 - CLSobel3x3 - CLSobel5x5 - CLSobel7x7 - CLTableLookup - CLThreshold - CLWarpAffine - CLWarpPerspective v21.02 Public major release - Various bug fixes. - Various optimisations. - Upgrade C++ standard to C++14 - Add macOS support - Add Armv8-R AArch64 architecture support - Add SVE/SVE2 support for: - NEScaleKernel - @ref NEActivationLayer - @ref NEArithmeticAddition - @ref NEBatchNormalizationLayerKernel - @ref cpu::kernels::CpuLogits1DSoftmaxKernel - @ref cpu::kernels::CpuLogits1DMaxKernel - @ref cpu::kernels::CpuElementwiseUnaryKernel - Remove padding from OpenCL kernels: - CLDirectConvolutionLayerKernel - @ref CLArgMinMaxLayerKernel - @ref CLPadLayerKernel - @ref CLROIAlignLayerKernel - @ref CLRangeKernel - CLScaleKernel - @ref CLSelectKernel - @ref CLBitwiseKernel - @ref opencl::kernels::ClFloorKernel - CLTransposeKernel - Deprecate functions in CLTuner: - add_lws_to_table - import_lws_table - lws_table - Remove functions: - NELocallyConnectedLayer / CLLocallyConnectedLayer - NEIm2Col - NECol2Im - NEGEMMInterleave4x4 - NEGEMMTranspose1xW - NEComputeAllAnchors / CLComputeAllAnchors - NEGEMMAssemblyDispatch - NEUpsampleLayer / CLUpsampleLayer - Remove kernels: - NEGEMMMatrixVectorMultiplyKernel - NELocallyConnectedMatrixMultiplyKernel / CLLocallyConnectedMatrixMultiplyKernel - NEUpsampleLayerKernel / CLUpsampleLayerKernel - Extend OpenCL tuner with workgroup batch size support - Experimental extension for the OpenCL tuner to tune the batches of work groups distribute to compute units - Add functionality to load the OpenCL GEMM heuristics at runtime - The GEMM heuristic file (MLGO) can be used to update the default GEMM heuristics available for OpenCL - Note: there might be performance regressions against v20.08 in Inception v3 using int8 data types on Arm Mali-G77 GPUs. Currently under investigation - Note: data-type decoupling is in progress and expiremental. Warning of unused symbols might be raised v20.11 Public major release - Various bug fixes. - Various optimisations. - Performance regressions can be noted when executing Depthwise Convolution on Arm® Neon™ with a depth multiplier > 1 for quantized data type. This is planned to be resolved in 21.02 release. - Added new data type QASYMM8_SIGNED support for @ref NEROIAlignLayer. - Added new data type S32 support for: - NEArithmeticSubtraction - NEArithmeticSubtractionKernel - @ref NEPixelWiseMultiplication - NEPixelWiseMultiplicationKernel - NEElementwiseDivision - NEDivisionOperationKernel - Interface change - Properly support softmax axis to have the same meaning as other major frameworks. That is, axis now defines the dimension on which Softmax/Logsoftmax is performed. E.g. for input of shape 4x5x6 and axis=1, softmax will be applied to 4x6=24 vectors of size 5. The supported value range of axis is [-rank, rank). This change applies to the following functions: - @ref NESoftmaxLayer - @ref NELogSoftmaxLayer - @ref CLSoftmaxLayer - @ref CLLogSoftmaxLayer - GCSoftmaxLayer - New OpenCL kernels / functions: - @ref CLGEMMLowpQuantizeDownInt32ScaleByFixedPointKernel - @ref CLLogicalNot - @ref CLLogicalAnd - @ref CLLogicalOr - New Arm® Neon™ kernels / functions: - @ref NELogicalNot - @ref NELogicalAnd - @ref NELogicalOr - Removed padding from Arm® Neon™ kernels: - NEComplexPixelWiseMultiplicationKernel - NENonMaximaSuppression3x3Kernel - @ref NERemapKernel - @ref NEGEMMInterleave4x4Kernel - NEDirectConvolutionLayerKernel - NEScaleKernel - NELocallyConnectedMatrixMultiplyKernel - @ref NEGEMMLowpOffsetContributionKernel - @ref NEGEMMTranspose1xWKernel - NEPoolingLayerKernel - NEConvolutionKernel - NEDepthwiseConvolutionLayerNativeKernel - @ref NEGEMMLowpMatrixMultiplyKernel - @ref NEGEMMMatrixMultiplyKernel - NEDirectConvolutionLayerOutputStageKernel - @ref NEReductionOperationKernel - @ref NEGEMMLowpMatrixAReductionKernel - @ref NEGEMMLowpMatrixBReductionKernel - Removed padding from OpenCL kernels: - CLBatchConcatenateLayerKernel - CLElementwiseOperationKernel - @ref CLBatchNormalizationLayerKernel - CLPoolingLayerKernel - @ref CLWinogradInputTransformKernel - @ref CLGEMMLowpMatrixMultiplyNativeKernel - @ref CLGEMMLowpMatrixAReductionKernel - @ref CLGEMMLowpMatrixBReductionKernel - @ref CLGEMMLowpOffsetContributionOutputStageKernel - @ref CLGEMMLowpOffsetContributionKernel - @ref CLWinogradOutputTransformKernel - @ref CLGEMMLowpMatrixMultiplyReshapedKernel - @ref CLFuseBatchNormalizationKernel - @ref CLDepthwiseConvolutionLayerNativeKernel - @ref CLDepthConvertLayerKernel - CLCopyKernel - @ref CLDepthwiseConvolutionLayer3x3NHWCKernel - CLActivationLayerKernel - @ref CLWinogradFilterTransformKernel - CLWidthConcatenateLayerKernel - CLWidthConcatenate4TensorsKernel - CLWidthConcatenate2TensorsKernel - CLLogits1DMaxShiftExpSumKernel - CLLogits1DNormKernel - CLHeightConcatenateLayerKernel - @ref CLGEMMMatrixMultiplyKernel - @ref CLGEMMLowpQuantizeDownInt32ScaleKernel - @ref CLGEMMLowpQuantizeDownInt32ScaleByFloatKernel - @ref CLGEMMLowpMatrixMultiplyReshapedOnlyRHSKernel - CLDepthConcatenateLayerKernel - @ref CLGEMMLowpQuantizeDownInt32ScaleByFixedPointKernel - Removed OpenCL kernels / functions: - CLGEMMLowpQuantizeDownInt32ToInt16ScaleByFixedPointKernel - CLGEMMLowpQuantizeDownInt32ToInt8ScaleByFixedPointKernel - CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPointKernel - Deprecated OpenCL kernels / functions (If a kernel is used only by the function that is being deprecated, the kernel is deprecated together): - CLLocallyConnectedLayer - CLLocallyConnectedMatrixMultiplyKernel - CLAbsoluteDifference - CLAbsoluteDifferenceKernel - CLAccumulate - CLAccumulateKernel - CLAccumulateSquared - CLAccumulateSquaredKernel - CLAccumulateWeighted - CLAccumulateWeightedKernel - CLAccumulateWeightedFP16Kernel - CLBox3x3 - CLBox3x3Kernel - CLBox3x3FP16Kernel - CLCannyEdge - CLChannelCombine - CLChannelCombineKernel - CLChannelExtract - CLChannelExtractKernel - CLColorConvert - CLColorConvertKernel - CLConvolution3x3 - CLConvolutionRectangle - CLConvolutionRectangleKernel - CLConvolutionSquare - CLConvolutionKernel - CLDerivative - CLDerivativeKernel - CLDilate - CLDilateKernel - CLEqualizeHistogram - CLErode - CLErodeKernel - CLFastCorners - CLFastCornersKernel - CLGaussian3x3 - CLGaussian3x3Kernel - CLGaussian5x5 - CLGaussian5x5HorKernel - CLGaussian5x5VertKernel - CLGaussianPyramid - CLGaussianPyramidHalf - CLGaussianPyramidOrb - CLHarrisCorners - CLHarrisScoreKernel - CLHarrisScoreFP16Kernel - CLHistogram - CLHistogramKernel - CLHOGOrientationBinningKernel - CLHOGBlockNormalizationKernel - CLHOGDetectorKernel - CLHOGNonMaximaSuppressionKernel - CLHOGDescriptor - CLHOGDetector - CLHOGGradient - CLHOGMultiDetection - CLHOGOrientationBinningKernel - CLHOGBlockNormalizationKernel - CLHOGDetectorKernel - CLIntegralImage - CLIntegralImageKernel - CLLaplacianReconstruct - CLLaplacianPyramid - CLMagnitude - CLMagnitudePhaseKernel - CLMedian3x3 - CLMedian3x3Kernel - CLMinMaxLocation - CLMinMaxLocationKernel - CLNonLinearFilter - CLNonLinearFilterKernel - CLNonMaximaSuppression3x3 - CLNonMaximaSuppression3x3FP16Kernel - CLNonMaximaSuppression3x3Kernel - CLOpticalFlow - CLPhase - CLRemap - CLRemapKernel - CLScharr3x3 - CLScharr3x3Kernel - CLSobel3x3 - CLSobel3x3Kernel - CLSobel5x5 - CLSobel5x5HorKernel - CLSobel5x5VertKernel - CLSobel7x7 - CLSobel7x7HorKernel - CLSobel7x7VertKernel - CLThreshold - CLThresholdKernel - CLWarpAffine - CLWarpAffineKernel - CLWarpPerspective - CLWarpPerspectiveKernel - Deprecated Arm® Neon™ kernels / functions (If a kernel is used only by the function that is being deprecated, the kernel is deprecated together): - NELocallyConnectedLayer - NELocallyConnectedMatrixMultiplyKernel - NEAbsoluteDifference - NEAbsoluteDifferenceKernel - NEAccumulate - NEAccumulateKernel - NEAccumulateSquared - NEAccumulateSquaredKernel - NEAccumulateWeighted - NEAccumulateWeightedKernel - NEAccumulateWeightedFP16Kernel - NEBox3x3 - NEBox3x3Kernel - NEBox3x3FP16Kernel - NECannyEdge - NEChannelCombine - NEChannelCombineKernel - NEChannelExtract - NEChannelExtractKernel - NEColorConvert - NEColorConvertKernel - NEConvolution3x3 - NEConvolutionRectangle - NEConvolutionRectangleKernel - NEConvolutionSquare - NEConvolutionKernel - NEDerivative - NEDerivativeKernel - NEDilate - NEDilateKernel - NEEqualizeHistogram - NEErode - NEErodeKernel - NEFastCorners - NEFastCornersKernel - NEGaussian3x3 - NEGaussian3x3Kernel - NEGaussian5x5 - NEGaussian5x5HorKernel - NEGaussian5x5VertKernel - NEGaussianPyramid - NEGaussianPyramidHalf - NEGaussianPyramidOrb - NEHarrisCorners - NEHarrisScoreKernel - NEHarrisScoreFP16Kernel - NEHistogram - NEHistogramKernel - NEHOGOrientationBinningKernel - NEHOGBlockNormalizationKernel - NEHOGDetectorKernel - NEHOGNonMaximaSuppressionKernel - NEHOGDescriptor - NEHOGDetector - NEHOGGradient - NEHOGMultiDetection - NEHOGOrientationBinningKernel - NEHOGBlockNormalizationKernel - NEHOGDetectorKernel - NEIntegralImage - NEIntegralImageKernel - NELaplacianReconstruct - NELaplacianPyramid - NEMagnitude - NEMagnitudePhaseKernel - NEMedian3x3 - NEMedian3x3Kernel - NEMinMaxLocation - NEMinMaxLocationKernel - NENonLinearFilter - NENonLinearFilterKernel - NENonMaximaSuppression3x3 - NENonMaximaSuppression3x3FP16Kernel - NENonMaximaSuppression3x3Kernel - NEOpticalFlow - NEPhase - NERemap - NERemapKernel - NEScharr3x3 - NEScharr3x3Kernel - NESobel3x3 - NESobel3x3Kernel - NESobel5x5 - NESobel5x5HorKernel - NESobel5x5VertKernel - NESobel7x7 - NESobel7x7HorKernel - NESobel7x7VertKernel - NEThreshold - NEThresholdKernel - NEWarpAffine - NEWarpAffineKernel - NEWarpPerspective - NEWarpPerspectiveKernel - Deprecated GLES kernels / functions (If a kernel is used only by the function that is being deprecated, the kernel is deprecated together): - GCAbsoluteDifference - GCActivationLayer - GCArithmeticAddition - GCBatchNormalizationLayer - GCConcatenateLayer - GCConvolutionLayer - GCDepthwiseConvolutionLayer - GCDirectConvolutionLayer - GCDropoutLayer - GCFillBorder - GCFullyConnectedLayer - GCGEMM - GCGEMMInterleave4x4 - GCGEMMTranspose1xW - GCNormalizationLayer - GCNormalizePlanarYUVLayer - GCPixelWiseMultiplication - GCPoolingLayer - GCScale - GCSoftmaxLayer - GCTensorShift - GCTranspose v20.08 Public major release - Various bug fixes. - Various optimisations. - Added new data type QASYMM8_SIGNED support for: - @ref CLArgMinMaxLayer - @ref CLArgMinMaxLayerKernel - Added new data type U8 support for: - @ref NECropKernel - CLCropKernel - Added aligh_corner support for nearest neighbor interpolation in: - NEScaleKernel - CLScaleKernel - New OpenCL kernels / functions: - @ref CLMaxUnpoolingLayerKernel - New Arm® Neon™ kernels / functions: - @ref NEMaxUnpoolingLayerKernel - New graph example: - graph_yolov3_output_detector - GEMMTuner improvements: - Added fp16 support - Output json files for easier integration - Enabled tuning for export_to_cl_image_rhs option for RHS tensors - More robust script for running benchmarks - Removed padding from: - NEPixelWiseMultiplicationKernel - NEHeightConcatenateLayerKernel - NEThresholdKernel - NEBatchConcatenateLayerKernel - NETransposeKernel - @ref NEBatchNormalizationLayerKernel - NEArithmeticSubtractionKernel - @ref NEBoundingBoxTransformKernel - NELogits1DMaxKernel - NELogits1DSoftmaxKernel - @ref NEROIPoolingLayerKernel - @ref NEROIAlignLayerKernel - NEYOLOLayerKernel - NEUpsampleLayerKernel - NEFloorKernel - NEWidthConcatenateLayerKernel - NEDepthConcatenateLayerKernel - @ref NENormalizationLayerKernel - @ref NEL2NormalizeLayerKernel - NEFillArrayKernel - @ref NEDepthConvertLayerKernel - @ref NERangeKernel - @ref NEPriorBoxLayer - Removed OpenCL kernels / functions: - CLGEMMLowpQuantizeDownInt32ToUint8Scale - CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFloat - Removed Arm® Neon™ kernels / functions: - NEGEMMLowpQuantizeDownInt32ToUint8Scale - NEGEMMMatrixAccumulateBiasesKernel - Deprecated functions / interfaces: - Non-descriptor based interfaces for NEThreshold, CLThreshold - Non-descriptor based interfaces for @ref NEScale, @ref CLScale and GCScale - In @ref NESoftmaxLayer, @ref NELogSoftmaxLayer, @ref CLSoftmaxLayer, @ref CLLogSoftmaxLayer and GCSoftmaxLayer : The default "axis" value for @ref CLSoftmaxLayer, @ref CLLogSoftmaxLayer and GCSoftmaxLayer is changed from 1 to 0. Only axis 0 is supported. The default "axis" value for @ref NESoftmaxLayer, @ref NELogSoftmaxLayer is changed from 1 to 0. Only axis 0 is supported. - The support for quantized data types has been removed from @ref CLLogSoftmaxLayer due to implementation complexity. - Removed padding requirement for the input (e.g. LHS of GEMM) and output in @ref CLGEMMMatrixMultiplyNativeKernel, @ref CLGEMMMatrixMultiplyReshapedKernel, @ref CLGEMMMatrixMultiplyReshapedOnlyRHSKernel and @ref CLIm2ColKernel (NHWC only) - This change allows to use @ref CLGEMMConvolutionLayer without extra padding for the input and output. - Only the weights/bias of @ref CLGEMMConvolutionLayer could require padding for the computation. - Only on Arm® Mali™ Midgard GPUs, @ref CLGEMMConvolutionLayer could require padding since @ref CLGEMMMatrixMultiplyKernel is called and currently requires padding. - Added support for exporting the OpenCL buffer object to the OpenCL image object in @ref CLGEMMMatrixMultiplyReshapedKernel and @ref CLGEMMMatrixMultiplyReshapedOnlyRHSKernel. - This support allows to export the OpenCL buffer used for the reshaped RHS matrix to the OpenCL image object. - The padding requirement for the OpenCL image object is considered into the @ref CLGEMMReshapeRHSMatrixKernel. - The reshaped RHS matrix stores the weights when GEMM is used to accelerate @ref CLGEMMConvolutionLayer. v20.05 Public major release - Various bug fixes. - Various optimisations. - Updated recommended NDK version to r18b. - Updated recommended gcc version to Linaro 6.3.1. - Added Bfloat16 type support - Added Bfloat16 support in: - @ref NEWeightsReshapeKernel - @ref NEConvolutionLayerReshapeWeights - @ref NEIm2ColKernel - NEIm2Col - @ref NEDepthConvertLayerKernel - @ref NEDepthConvertLayer - @ref NEGEMMConvolutionLayer - 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 NEScale - NEScaleKernel - 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 - @ref CPPUpsample - @ref CPPUpsampleKernel - New OpenCL kernels / functions: - @ref CLQLSTMLayer - @ref CLQLSTMLayerNormalizationKernel - New Arm® Neon™ kernels / functions: - @ref NEQLSTMLayer - @ref NEQLSTMLayerNormalizationKernel - Added HARD_SWISH support in: - CLActivationLayerKernel - NEActivationLayerKernel - Deprecated OpenCL kernels / functions: - CLGEMMLowpQuantizeDownInt32ToUint8Scale - CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFloat - Deprecated Arm® Neon™ kernels / functions: - NEGEMMLowpQuantizeDownInt32ToUint8Scale - Removed CPP kernels / functions: - CPPFlipWeightsKernel - Removed PoolingLayerInfo constructors without Data Layout. - Removed CLDepthwiseConvolutionLayer3x3 - Removed NEDepthwiseConvolutionLayerOptimized - Added support for Winograd 3x3,4x4 on Arm® Neon™ FP16: - @ref NEWinogradConvolutionLayer - @ref NEWinogradLayerTransformInputKernel - @ref NEWinogradLayerTransformOutputKernel - @ref NEWinogradLayerTransformWeightsKernel - Added CLCompileContext - Added Arm® Neon™ GEMM kernel with 2D window support v20.02.1 Maintenance release - Added Android-NN build script. v20.02 Public major release - Various bug fixes. - Various optimisations. - Added new data type QASYMM8_SIGNED support for: - @ref CLDepthwiseConvolutionLayer - CLDepthwiseConvolutionLayer3x3 - @ref CLGEMMConvolutionLayer - @ref CLGEMMLowpMatrixMultiplyCore - @ref CLGEMMLowpMatrixMultiplyReshapedOnlyRHSKernel - @ref CLGEMMLowpMatrixMultiplyNativeKernel - @ref NEActivationLayer - NEComparisonOperationKernel - @ref NEConvolutionLayer - @ref NEDepthwiseConvolutionLayer - NEDepthwiseConvolutionLayer3x3Kernel - NEDirectConvolutionLayerOutputStageKernel - @ref NEElementwiseComparison - @ref NEElementwiseMax - @ref NEElementwiseMin - @ref NEElementwiseSquaredDiff - @ref NEFullyConnectedLayer - NEGEMMMatrixVectorMultiplyKernel - @ref NEPixelWiseMultiplication - @ref NEPoolingLayer - @ref NEPReluLayer - Added support for QSYMM8_PER_CHANNEL in: - NEDepthwiseConvolutionLayer3x3Kernel - Added support for split sizes in: - @ref CLSplit - @ref NESplit - New OpenCL kernels / functions: - @ref CLFill - CLGEMMLowpQuantizeDownInt32ToInt8ScaleByFixedPointKernel / @ref CLGEMMLowpQuantizeDownInt32ToInt8ScaleByFixedPoint - New Arm® Neon™ kernels / functions: - @ref NEFill - @ref NEGEMMLowpQuantizeDownInt32ToInt8ScaleByFixedPointKernel / @ref NEGEMMLowpQuantizeDownInt32ToInt8ScaleByFixedPoint - Deprecated Arm® Neon™ functions / interfaces: - CLDepthwiseConvolutionLayer3x3 - NEDepthwiseConvolutionLayerOptimized - PoolingLayerInfo constructors without Data Layout. - Added support for quantization with multiplier greater than 1 on Arm® Neon™ and CL. - Added support for quantized inputs of type QASYMM8_SIGNED and QASYMM8 to @ref CLQuantizationLayer. - Added the ability to build bootcode for bare metal. - Added support for generating synthetic QASYMM8 graphs. - Added support for F16 datatype in VGG16. - Removed pre-built binaries for GLES. v19.11.1 Public maintenance release - Fix offset calculation in NEReductionOperationKernel. - Fix data layout in NEScaleKernel for nhwc. - Retain configuration step data layout to avoid side-effects. - Perform sqrt in double domain for L2 pooling. - Fix output shape calculation for Reduce Mean - Restrict cases where optimized NEPadLayer runs. v19.11 Public major release - Various bug fixes. - Various optimisations. - Updated recommended NDK version to r17c. - Deprecated OpenCL kernels / functions: - CLDepthwiseConvolutionLayerReshapeWeightsGenericKernel - CLDepthwiseIm2ColKernel - CLDepthwiseSeparableConvolutionLayer - CLDepthwiseVectorToTensorKernel - CLDirectConvolutionLayerOutputStageKernel - Deprecated Arm® Neon™ kernels / functions: - NEDepthwiseWeightsReshapeKernel - NEDepthwiseIm2ColKernel - NEDepthwiseSeparableConvolutionLayer - NEDepthwiseVectorToTensorKernel - NEDepthwiseConvolutionLayer3x3 - New OpenCL kernels / functions: - @ref CLInstanceNormalizationLayerKernel / @ref CLInstanceNormalizationLayer - @ref CLDepthwiseConvolutionLayerNativeKernel to replace the old generic depthwise convolution (see Deprecated OpenCL kernels / functions) - @ref CLLogSoftmaxLayer - New Arm® Neon™ kernels / functions: - @ref NEBoundingBoxTransformKernel / @ref NEBoundingBoxTransform - @ref NEComputeAllAnchorsKernel / NEComputeAllAnchors - @ref NEDetectionPostProcessLayer - @ref NEGenerateProposalsLayer - @ref NEInstanceNormalizationLayerKernel / @ref NEInstanceNormalizationLayer - @ref NELogSoftmaxLayer - @ref NEROIAlignLayerKernel / @ref NEROIAlignLayer - Added QASYMM8 support for: - @ref CLGenerateProposalsLayer - @ref CLROIAlignLayer - @ref CPPBoxWithNonMaximaSuppressionLimit - Added QASYMM16 support for: - @ref CLBoundingBoxTransform - Added FP16 support for: - @ref CLGEMMMatrixMultiplyReshapedKernel - Added new data type QASYMM8_PER_CHANNEL support for: - CLDequantizationLayer - @ref NEDequantizationLayer - Added new data type QSYMM8_PER_CHANNEL support for: - @ref CLConvolutionLayer - @ref NEConvolutionLayer - @ref CLDepthwiseConvolutionLayer - @ref NEDepthwiseConvolutionLayer - Added FP16 mixed-precision support for: - @ref CLGEMMMatrixMultiplyReshapedKernel - CLPoolingLayerKernel - Added FP32 and FP16 ELU activation for: - @ref CLActivationLayer - @ref NEActivationLayer - Added asymmetric padding support for: - @ref CLDirectDeconvolutionLayer - @ref CLGEMMDeconvolutionLayer - @ref NEDeconvolutionLayer - Added SYMMETRIC and REFLECT modes for @ref CLPadLayerKernel / @ref CLPadLayer. - Replaced the calls to NECopyKernel and NEMemsetKernel with @ref NEPadLayer in @ref NEGenerateProposalsLayer. - Replaced the calls to CLCopyKernel and CLMemsetKernel with @ref CLPadLayer in @ref CLGenerateProposalsLayer. - Improved performance for CL Inception V3 - FP16. - Improved accuracy for CL Inception V3 - FP16 by enabling FP32 accumulator (mixed-precision). - Improved Arm® Neon™ performance by enabling fusing batch normalization with convolution and depth-wise convolution layer. - Improved Arm® Neon™ performance for MobileNet-SSD by improving the output detection performance. - Optimized @ref CLPadLayer. - Optimized CL generic depthwise convolution layer by introducing @ref CLDepthwiseConvolutionLayerNativeKernel. - Reduced memory consumption by implementing weights sharing. v19.08.1 Public maintenance release - Fix offset calculation in NEReductionOperationKernel. - Fix data layout in NEScaleKernel for nhwc. - Retain configuration step data layout to avoid side-effects. - Perform sqrt in double domain for L2 pooling. - Fix output shape calculation for Reduce Mean - Fix broadcast CLPixelwiseMultiplication with 5D tensors v19.08 Public major release - Various bug fixes. - Various optimisations. - Deprecated Arm® Neon™ functions - NEDepthConcatenateLayer - NEWidthConcatenateLayer - Deprecated OpenCL kernels / functions - CLDepthConcatenateLayer - CLGEMMInterleave4x4Kernel / CLGEMMInterleave4x4 - CLGEMMTranspose1xWKernel / CLGEMMTranspose1xW - CLWidthConcatenateLayer - New Arm® Neon™ kernels / functions: - @ref NEAbsLayer - @ref NECast - @ref NEElementwisePower - @ref NELogLayer - @ref NELSTMLayerQuantized - @ref NENegLayer - @ref NEPReluLayer - @ref NESinLayer - NEBatchConcatenateLayerKernel - @ref NEDepthToSpaceLayerKernel / @ref NEDepthToSpaceLayer - NEDepthwiseConvolutionLayerNativeKernel - @ref NEGEMMLowpQuantizeDownInt32ToInt16ScaleByFixedPointKernel - @ref NEMeanStdDevNormalizationKernel / @ref NEMeanStdDevNormalizationLayer - @ref NESpaceToDepthLayerKernel / @ref NESpaceToDepthLayer - New OpenCL kernels / functions: - @ref CLAbsLayer - @ref CLElementwisePower - @ref CLLogLayer - @ref CLLSTMLayerQuantized - @ref CLNegLayer - @ref CLPReluLayer - @ref CLSinLayer - CLBatchConcatenateLayerKernel - @ref CLDepthToSpaceLayerKernel / @ref CLDepthToSpaceLayer - @ref CLGEMMLowpMatrixMultiplyNativeKernel - CLGEMMLowpQuantizeDownInt32ToInt16ScaleByFixedPointKernel - @ref CLGEMMMatrixMultiplyNativeKernel - CLMeanStdDevNormalizationKernel /CLMeanStdDevNormalizationLayer - @ref CLSpaceToDepthLayerKernel / @ref CLSpaceToDepthLayer - New examples: - neon_opticalflow - cl_cache - neon_permute - Added support for FP16 in @ref NEDeconvolutionLayer - Added support for FP16 in @ref CLDeconvolutionLayer - Added support for REDUCE_MIN and REDUCE_MAX in @ref ReductionOperation - Enable the fusion of batch normalization with convolution and depthwise convolution layer for FP32 in the graph API (OpenCL only) - Added support for fusing activation function and broadcast addition with the matrix multiplication for FP32 (OpenCL only) - Re-factored the depthwise convolution layer kernel on Arm® Neon™ for generic cases - Added an optimized depthwise convolution layer kernel for 5x5 filters (Neon only) - Added support to enable OpenCL kernel cache. Added example showing how to load the prebuilt OpenCL kernels from a binary cache file - Altered @ref QuantizationInfo interface to support per-channel quantization. - The CLDepthwiseConvolutionLayer3x3 will be included by @ref CLDepthwiseConvolutionLayer to accommodate for future optimizations. - The NEDepthwiseConvolutionLayerOptimized will be included by @ref NEDepthwiseConvolutionLayer to accommodate for future optimizations. - Removed inner_border_right and inner_border_top parameters from @ref CLDeconvolutionLayer interface - Removed inner_border_right and inner_border_top parameters from @ref NEDeconvolutionLayer interface - Optimized the Arm® Neon™ assembly kernel for GEMMLowp. The new implementation fuses the output stage and quantization with the matrix multiplication kernel v19.05 Public major release - Various bug fixes. - Various optimisations. - New Arm® Neon™ kernels / functions: - @ref NEBatchToSpaceLayerKernel / @ref NEBatchToSpaceLayer - NEComplexPixelWiseMultiplicationKernel / @ref NEComplexPixelWiseMultiplication - @ref NECropKernel / @ref NECropResize - NEDepthwiseConvolutionAssemblyDispatch - @ref NEFFTDigitReverseKernel - @ref NEFFTRadixStageKernel - @ref NEFFTScaleKernel - @ref NEGEMMLowpOffsetContributionOutputStageKernel - NEHeightConcatenateLayerKernel - @ref NESpaceToBatchLayerKernel / @ref NESpaceToBatchLayer - @ref NEFFT1D - @ref NEFFT2D - @ref NEFFTConvolutionLayer - New OpenCL kernels / functions: - CLComplexPixelWiseMultiplicationKernel / @ref CLComplexPixelWiseMultiplication - CLCropKernel / @ref CLCropResize - @ref CLDeconvolutionReshapeOutputKernel - @ref CLFFTDigitReverseKernel - @ref CLFFTRadixStageKernel - @ref CLFFTScaleKernel - @ref CLGEMMLowpMatrixMultiplyReshapedOnlyRHSKernel - @ref CLGEMMMatrixMultiplyReshapedOnlyRHSKernel - CLHeightConcatenateLayerKernel - @ref CLDirectDeconvolutionLayer - @ref CLFFT1D - @ref CLFFT2D - @ref CLFFTConvolutionLayer - @ref CLGEMMDeconvolutionLayer - New OpenGLES kernels / functions: - GCConcatenateLayer - Deprecated functions/interfaces - GCDepthConcatenateLayer - NEWidthConcatenateLayer - NEDepthConcatenateLayer - CLWidthConcatenateLayer - CLDepthConcatenateLayer - CLGEMMInterleave4x4 - CLGEMMTranspose1xW - Support different quantization info in CLConcatLayer. - Add checks on different input/output quantization info were not supported. - Tensors have different quantization information. - Add FP16 support checks. - Fix output quantization CLDeptwiseConv3x3 when activation is fused. - New graph examples: - graph_convolution - graph_fully_connected - graph_depthwise_convolution - Deepspeech v0.4.1 - Add support for QASYMM8 in NEArithmeticSubtractionKernel. - Add support for QASYMM8 in NEPixelWiseMultiplicationKernel. - Add support for QASYMM8 NEDeconvolution. - Add support for DequantizationLayer for Neon/CL. - Add support for dilation in CLDepthwiseConvolution. - Fuse offset contribution with the output stage when we use NEGEMMLowpMatrixMultiplyCore. - Optimize CLDeconvolution. - Add StackLayer to the graph API. - Add support for "reflect" padding mode in NEPad. - Winograd 7x7 NHWC on OpenCL. - Rework CL ML layers to run exclusively on CL. - Support different quantization info in PoolingLayer. - Implement and test import memory interfaces. - Added new tests and removed old ones. - Various clang-tidy fixes. v19.02 Public major release - Various bug fixes. - Various optimisations. - New Arm® Neon™ kernels / functions: - @ref NETileKernel / @ref NETile - @ref NEFuseBatchNormalizationKernel / @ref NEFuseBatchNormalization - NEElementwiseOperationKernel - @ref NEElementwiseMax - @ref NEElementwiseMin - @ref NEElementwiseSquaredDiff - @ref NESelectKernel / @ref NESelect - @ref NESplit - @ref NESlice - @ref NEUnstack - @ref NEStridedSliceKernel / @ref NEStridedSlice - NEElementwiseUnaryKernel - @ref NERsqrtLayer - @ref NEExpLayer - @ref NEReverseKernel / @ref NEReverse - @ref NEArgMinMaxLayer - @ref NEStackLayerKernel / @ref NEStackLayer - @ref NERangeKernel / @ref NERange - @ref NEPadLayer - NEMemsetKernel - @ref NEGatherKernel / @ref NEGather - @ref NEElementwiseComparison - @ref NEElementwiseComparisonStatic - NEComparisonOperationKernel - @ref NEElementwiseDivision - New OpenCL kernels / functions: - @ref CLSelectKernel / @ref CLSelect - @ref CLTileKernel / @ref CLTile - @ref CLComparisonKernel / @ref CLComparison - @ref CLArgMinMaxLayer - @ref CLElementwiseMax - @ref CLElementwiseMin - @ref CLElementwiseSquaredDiff - @ref CLStackLayerKernel / @ref CLStackLayer - @ref CLReverse / @ref CLReverseKernel - @ref CLRsqrtLayer - @ref CLExpLayer - CLElementWiseUnaryLayerKernel - @ref CLGEMMReshapeLHSMatrixKernel - @ref CLGEMMReshapeRHSMatrixKernel - @ref CLGEMMMatrixMultiplyReshapedKernel - @ref CLRangeKernel / @ref CLRange - @ref CLUnstack - @ref CLGatherKernel / @ref CLGather - @ref CLGEMMLowpMatrixMultiplyReshapedKernel - New CPP kernels / functions: - @ref CPPDetectionOutputLayer - @ref CPPTopKV / @ref CPPTopKVKernel - Added new examples: - graph_ssd_mobilenet.cpp - graph_mobilenet_v2.cpp - graph_resnet12.cpp - graph_srcnn955.cpp - graph_vgg_vdsr.cpp - graph_inception_resnet_v1.cpp - Add 4D tensors support to - @ref NESoftmaxLayer - Fused activation in @ref CLWinogradConvolutionLayer - Extented @ref NEPermute to support more cases - Added Neon/SVE GEMM Hybrid kernels - Added u8 and s8 hybrid assembly kernels - Introduced GEMM strategy name in NEGEMMAssemblyWrapper - Improved @ref CLTuner - Fused the bias addition within @ref CLGEMM - Added support for QASYMM8 LOGISTIC activation in @ref NEActivationLayer - Added NHWC data layout support to: - @ref NEScale for F16 - @ref CLNormalizationLayer IN_MAP_2D for FP32/FP16 - @ref NEL2NormalizeLayer for FP32/FP16 - @ref NENormalizationLayer IN_MAP_2D for FP32/FP16 - @ref CLROIAlignLayer - @ref CLGenerateProposalsLayer - Added QASYMM8 support to the following kernels: - NEArithmeticAdditionKernel - @ref NEScale - Added new tests and improved validation and benchmarking suites. - Deprecated functions/interfaces - Usage of inner_border_right and inner_border_top has been deprecated in @ref CLDeconvolutionLayer and @ref NEDeconvolutionLayer v18.11 Public major release - Various bug fixes. - Various optimisations. - New Arm® Neon™ kernels / functions: - @ref NEChannelShuffleLayer / @ref NEChannelShuffleLayerKernel - @ref NEReduceMean - @ref NEReorgLayer / @ref NEReorgLayerKernel - @ref NEPriorBoxLayer / @ref NEPriorBoxLayerKernel - NEUpsampleLayer / NEUpsampleLayerKernel - NEYOLOLayer / NEYOLOLayerKernel - New OpenCL kernels / functions: - @ref CLBatchToSpaceLayer / @ref CLBatchToSpaceLayerKernel - @ref CLBoundingBoxTransform / @ref CLBoundingBoxTransformKernel - @ref CLComputeAllAnchorsKernel - @ref CLGenerateProposalsLayer - @ref CLNormalizePlanarYUVLayer / @ref CLNormalizePlanarYUVLayerKernel - @ref CLReorgLayer / @ref CLReorgLayerKernel - @ref CLSpaceToBatchLayer / @ref CLSpaceToBatchLayerKernel - @ref CLPadLayer - @ref CLReduceMean - @ref CLPriorBoxLayer / @ref CLPriorBoxLayerKernel - @ref CLROIAlignLayer / @ref CLROIAlignLayerKernel - @ref CLSlice - @ref CLSplit - @ref CLStridedSlice / @ref CLStridedSliceKernel - CLUpsampleLayer / CLUpsampleLayerKernel - CLYOLOLayer / CLYOLOLayerKernel - New CPP kernels / functions: - @ref CPPBoxWithNonMaximaSuppressionLimit / @ref CPPBoxWithNonMaximaSuppressionLimitKernel - Added the validate method in: - @ref NEDepthConvertLayer - @ref NEFloor / @ref CLFloor - @ref NEGEMMMatrixAdditionKernel - @ref NEReshapeLayer / @ref CLReshapeLayer - @ref CLScale - Added new examples: - graph_shufflenet.cpp - graph_yolov3.cpp - Added documentation for add a new function or kernel. - Improved doxygen documentation adding a list of the existing functions. - Add 4D tensors support to - CLWidthConcatenateLayer - CLFlattenLayer - @ref CLSoftmaxLayer - Add dot product support for @ref CLDepthwiseConvolutionLayer3x3NHWCKernel non-unit stride - Add SVE support - Fused batch normalization into convolution layer weights in @ref CLFuseBatchNormalization - Fuses activation in @ref CLDepthwiseConvolutionLayer3x3NCHWKernel, @ref CLDepthwiseConvolutionLayer3x3NHWCKernel and @ref NEGEMMConvolutionLayer - Added NHWC data layout support to: - @ref CLChannelShuffleLayer - @ref CLDeconvolutionLayer - @ref CLL2NormalizeLayer - Added QASYMM8 support to the following kernels: - CLScaleKernel - NEDepthwiseConvolutionLayer3x3Kernel - CLPixelWiseMultiplicationKernel - Added FP16 support to the following kernels: - @ref CLDepthwiseConvolutionLayer3x3NHWCKernel - NEDepthwiseConvolutionLayer3x3Kernel - @ref CLNormalizePlanarYUVLayerKernel - @ref CLWinogradConvolutionLayer (5x5 kernel) - More tests added to both validation and benchmarking suites. v18.08 Public major release - Various bug fixes. - Various optimisations. - Updated recommended NDK version to r17b. - Removed support for QS8/QS16 data types. - Added support for grouped convolution in @ref CLConvolutionLayer. - Added NHWC data layout support to: - NEDepthConcatenateLayer / CLDepthConcatenateLayer - @ref NEWinogradConvolutionLayer / @ref CLWinogradConvolutionLayer - @ref CLDepthwiseConvolutionLayer - @ref CLDirectConvolutionLayer - @ref CLConvolutionLayer - @ref CLScale - @ref CLIm2ColKernel - New Arm® Neon™ kernels / functions: - @ref NERNNLayer - New OpenCL kernels / functions: - @ref CLArithmeticDivision - Introduced prepare() stage support in the graph API for GLES. - Added support for memory reusage when trying to allocate smaller CLTensors. - Enabled NHWC execution on graph examples. - Added JPEG accessor for validation purposes. - Added validate methods to some kernels / functions. v18.05 Public major release - Various bug fixes. - Various optimisations. - Major redesign in the interface for the neon kernels implemented in assembly. - Removed arm_compute::NEGEMMLowpAArch64A53Kernel / arm_compute::NEGEMMLowpAArch64Kernel / arm_compute::NEGEMMLowpAArch64V8P4Kernel / arm_compute::NEGEMMInterleavedBlockedKernel / arm_compute::NEGEMMLowpAssemblyMatrixMultiplyCore / arm_compute::NEHGEMMAArch64FP16Kernel - Added NEGEMMAssemblyWrapper and AssemblyKernelGlue which are used to execute assembly kernels in neon functions. - Minor changes to the CPUInfo type to make it compatible with the new assembly gemm interface. - Moved neon assembly kernels to the folder src/core/Neon/kernels/arm_gemm. - Improved doxygen documentation. - Improved memory management for layer's transitions. - Added support for NHWC data layout in tensors. - Added NHWC data layout support to: - @ref NEGEMMConvolutionLayer - @ref NEDirectConvolutionLayer - @ref NEPoolingLayer / @ref CLPoolingLayer - @ref NEBatchNormalizationLayer / @ref CLBatchNormalizationLayer - @ref NEDepthwiseConvolutionLayer - @ref NEScale - NEIm2Col - Added support for dilated convolutions in @ref NEConvolutionLayer and @ref CLConvolutionLayer. - New OpenCL kernels / functions: - @ref CLChannelShuffleLayer / @ref CLChannelShuffleLayerKernel - CLConvertFullyConnectedWeightsKernel / @ref CLConvertFullyConnectedWeights - @ref CLCopy / CLCopyKernel - @ref CLLSTMLayer - @ref CLRNNLayer - CLWidthConcatenateLayer / CLWidthConcatenateLayerKernel - @ref CLWinogradFilterTransformKernel / @ref CLWinogradInputTransformKernel / @ref CLWinogradConvolutionLayer - @ref CLWinogradInputTransformKernel / @ref CLWinogradInputTransform - New Arm® Neon™ kernels / functions: - NEConvertFullyConnectedWeightsKernel / @ref NEConvertFullyConnectedWeights. - Created the validate method in @ref CLDepthwiseConvolutionLayer. - Beta and gamma are no longer mandatory arguments in @ref NEBatchNormalizationLayer and @ref CLBatchNormalizationLayer. - Added depth multiplier support in @ref NEDepthwiseConvolutionLayer and @ref CLDepthwiseConvolutionLayer. - Added broadcast multiply support in @ref NEPixelWiseMultiplication / NEPixelWiseMultiplicationKernel. - Port mobilenet example to NHWC data layout. - Enabled Winograd method in @ref CLConvolutionLayer. - Renamed NEWinogradLayer to @ref NEWinogradConvolutionLayer. - Updated @ref NEWinogradConvolutionLayer to use highly optimised assembly kernels in src/core/Neon/kernels/arm_gemm. - Added memory manager support in GLES functions. - Major refactoring of the graph API. - Added GLES backend in the graph API. - Added support for the memory manager in the graph API. - Enabled Winograd Convolution method in the graph API. - Added support for grouped convolutions in the graph API. - Replaced NEDeconvolutionLayerUpsampleKernel with NEScaleKernel in @ref NEDeconvolutionLayer. - Added fast maths flag in @ref CLConvolutionLayer. - Added new tests and benchmarks in validation and benchmark frameworks - Merge Activation layer with Convolution Layer (Neon. CL, GLES) - Added support to OpenCL 2.0 SVM - Added support to import memory in OpenCL tensors. - Added the prepare() method to perform any one off pre-processing before running the function. - Added new examples: - graph_inception_v4.cpp - graph_resnext50.cpp - Added memory measurement instrument for CL. v18.03 Public maintenance release - Various bug fixes. - Fixed bug in @ref NEActivationLayer - Fix in @ref CLTuner when using batches. - Updated recommended NDK version to r16b (And fixed warnings). - Fixed bug in validation code. - Added Inception v4 graph example. - Renamed NEWinogradLayer.cpp to @ref NEWinogradConvolutionLayer v18.02 Public major release - Various Arm® Neon™ / OpenCL / GLES optimisations. - Various bug fixes. - Changed default number of threads on big LITTLE systems. - Refactored examples and added: - graph_mobilenet_qassym8 - graph_resnet - graph_squeezenet_v1_1 - Renamed @ref CLConvolutionLayer into @ref CLGEMMConvolutionLayer and created a new @ref CLConvolutionLayer to select the fastest convolution method. - Renamed @ref NEConvolutionLayer into @ref NEGEMMConvolutionLayer and created a new @ref NEConvolutionLayer to select the fastest convolution method. - Added in place support to: - @ref CLActivationLayer - @ref CLBatchNormalizationLayer - Added QASYMM8 support to: - @ref CLActivationLayer - @ref CLDepthwiseConvolutionLayer - @ref NEDepthwiseConvolutionLayer - @ref NESoftmaxLayer - Added FP16 support to: - CLDepthwiseConvolutionLayer3x3 - @ref CLDepthwiseConvolutionLayer - Added broadcasting support to NEArithmeticAddition / @ref CLArithmeticAddition / @ref CLPixelWiseMultiplication - Added fused batched normalization and activation to @ref CLBatchNormalizationLayer and @ref NEBatchNormalizationLayer - Added support for non-square pooling to @ref NEPoolingLayer and @ref CLPoolingLayer - New OpenCL kernels / functions: - CLDirectConvolutionLayerOutputStageKernel - New Arm® Neon™ kernels / functions - Added name() method to all kernels. - Added support for Winograd 5x5. - NEPermuteKernel / @ref NEPermute - @ref NEWinogradLayerTransformInputKernel / NEWinogradLayer - @ref NEWinogradLayerTransformOutputKernel / NEWinogradLayer - @ref NEWinogradLayerTransformWeightsKernel / NEWinogradLayer - Renamed NEWinogradLayerKernel into NEWinogradLayerBatchedGEMMKernel - New GLES kernels / functions: - GCTensorShiftKernel / GCTensorShift v18.01 Public maintenance release - Various bug fixes - Added some of the missing validate() methods - Added @ref CLDeconvolutionLayerUpsampleKernel / @ref CLDeconvolutionLayer @ref CLDeconvolutionLayerUpsample - Added CLPermuteKernel / @ref CLPermute - Added method to clean the programs cache in the CL Kernel library. - Added GCArithmeticAdditionKernel / GCArithmeticAddition - Added GCDepthwiseConvolutionLayer3x3Kernel / GCDepthwiseConvolutionLayer3x3 - Added GCNormalizePlanarYUVLayerKernel / GCNormalizePlanarYUVLayer - Added GCScaleKernel / GCScale - Added GCWeightsReshapeKernel / GCConvolutionLayer - Added FP16 support to the following GLES compute kernels: - GCCol2ImKernel - GCGEMMInterleave4x4Kernel - GCGEMMTranspose1xWKernel - GCIm2ColKernel - Refactored Arm® Neon™ Winograd (NEWinogradLayerKernel) - Added NEDirectConvolutionLayerOutputStageKernel - Added QASYMM8 support to the following Arm® Neon™ kernels: - NEDepthwiseConvolutionLayer3x3Kernel - @ref NEFillBorderKernel - NEPoolingLayerKernel - Added new examples: - graph_cl_mobilenet_qasymm8.cpp - graph_inception_v3.cpp - gc_dc.cpp - More tests added to both validation and benchmarking suites. v17.12 Public major release - Most machine learning functions on OpenCL support the new data type QASYMM8 - Introduced logging interface - Introduced opencl timer - Reworked GEMMLowp interface - Added new Arm® Neon™ assembly kernels for GEMMLowp, SGEMM and HGEMM - Added validation method for most Machine Learning kernels / functions - Added new graph examples such as googlenet, mobilenet, squeezenet, vgg16 and vgg19 - Added sgemm example for OpenCL - Added absolute difference example for GLES compute - Added new tests and benchmarks in validation and benchmark frameworks - Added new kernels / functions for GLES compute - New OpenGL ES kernels / functions - GCAbsoluteDifferenceKernel / GCAbsoluteDifference - GCActivationLayerKernel / GCActivationLayer - GCBatchNormalizationLayerKernel / GCBatchNormalizationLayer - GCCol2ImKernel - GCDepthConcatenateLayerKernel / GCDepthConcatenateLayer - GCDirectConvolutionLayerKernel / GCDirectConvolutionLayer - GCDropoutLayerKernel / GCDropoutLayer - GCFillBorderKernel / GCFillBorder - GCGEMMInterleave4x4Kernel / GCGEMMInterleave4x4 - GCGEMMMatrixAccumulateBiasesKernel / GCGEMMMatrixAdditionKernel / GCGEMMMatrixMultiplyKernel / GCGEMM - GCGEMMTranspose1xWKernel / GCGEMMTranspose1xW - GCIm2ColKernel - GCNormalizationLayerKernel / GCNormalizationLayer - GCPixelWiseMultiplicationKernel / GCPixelWiseMultiplication - GCPoolingLayerKernel / GCPoolingLayer - GCLogits1DMaxKernel / GCLogits1DShiftExpSumKernel / GCLogits1DNormKernel / GCSoftmaxLayer - GCTransposeKernel / GCTranspose - New Arm® Neon™ kernels / functions - arm_compute::NEGEMMLowpAArch64A53Kernel / arm_compute::NEGEMMLowpAArch64Kernel / arm_compute::NEGEMMLowpAArch64V8P4Kernel / arm_compute::NEGEMMInterleavedBlockedKernel / arm_compute::NEGEMMLowpAssemblyMatrixMultiplyCore - arm_compute::NEHGEMMAArch64FP16Kernel - NEDepthwiseConvolutionLayer3x3Kernel / NEDepthwiseIm2ColKernel / NEGEMMMatrixVectorMultiplyKernel / NEDepthwiseVectorToTensorKernel / @ref NEDepthwiseConvolutionLayer - @ref NEGEMMLowpOffsetContributionKernel / @ref NEGEMMLowpMatrixAReductionKernel / @ref NEGEMMLowpMatrixBReductionKernel / @ref NEGEMMLowpMatrixMultiplyCore - @ref NEGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPointKernel / @ref NEGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPoint - NEWinogradLayer / NEWinogradLayerKernel - New OpenCL kernels / functions - @ref CLGEMMLowpOffsetContributionKernel / @ref CLGEMMLowpMatrixAReductionKernel / @ref CLGEMMLowpMatrixBReductionKernel / @ref CLGEMMLowpMatrixMultiplyCore - CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPointKernel / @ref CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPoint - New graph nodes for Arm® Neon™ and OpenCL - graph::BranchLayer - graph::DepthConvertLayer - graph::DepthwiseConvolutionLayer - graph::DequantizationLayer - graph::FlattenLayer - graph::QuantizationLayer - graph::ReshapeLayer v17.10 Public maintenance release - Bug fixes: - Check the maximum local workgroup size supported by OpenCL devices - Minor documentation updates (Fixed instructions to build the examples) - Introduced a graph::GraphContext - Added a few new Graph nodes, support for branches and grouping. - Automatically enable cl_printf in debug builds - Fixed bare metal builds for armv7a - Added AlexNet and cartoon effect examples - Fixed library builds: libraries are no longer built as supersets of each other.(It means application using the Runtime part of the library now need to link against both libarm_compute_core and libarm_compute) v17.09 Public major release - Experimental Graph support: initial implementation of a simple stream API to easily chain machine learning layers. - Memory Manager (@ref BlobLifetimeManager, @ref BlobMemoryPool, @ref ILifetimeManager, @ref IMemoryGroup, @ref IMemoryManager, @ref IMemoryPool, @ref IPoolManager, @ref MemoryManagerOnDemand, @ref PoolManager) - New validation and benchmark frameworks (Boost and Google frameworks replaced by homemade framework). - Most machine learning functions support both fixed point 8 and 16 bit (QS8, QS16) for both Arm® Neon™ and OpenCL. - New Arm® Neon™ kernels / functions: - arm_compute::NEGEMMAssemblyBaseKernel arm_compute::NEGEMMAArch64Kernel - NEDequantizationLayerKernel / @ref NEDequantizationLayer - NEFloorKernel / @ref NEFloor - @ref NEL2NormalizeLayerKernel / @ref NEL2NormalizeLayer - NEQuantizationLayerKernel @ref NEMinMaxLayerKernel / @ref NEQuantizationLayer - @ref NEROIPoolingLayerKernel / @ref NEROIPoolingLayer - @ref NEReductionOperationKernel / @ref NEReductionOperation - NEReshapeLayerKernel / @ref NEReshapeLayer - New OpenCL kernels / functions: - @ref CLDepthwiseConvolutionLayer3x3NCHWKernel @ref CLDepthwiseConvolutionLayer3x3NHWCKernel CLDepthwiseIm2ColKernel CLDepthwiseVectorToTensorKernel CLDepthwiseWeightsReshapeKernel / CLDepthwiseConvolutionLayer3x3 @ref CLDepthwiseConvolutionLayer CLDepthwiseSeparableConvolutionLayer - CLDequantizationLayerKernel / CLDequantizationLayer - CLDirectConvolutionLayerKernel / @ref CLDirectConvolutionLayer - CLFlattenLayer - CLFloorKernel / @ref CLFloor - CLGEMMTranspose1xW - CLGEMMMatrixVectorMultiplyKernel - @ref CLL2NormalizeLayerKernel / @ref CLL2NormalizeLayer - CLQuantizationLayerKernel @ref CLMinMaxLayerKernel / @ref CLQuantizationLayer - @ref CLROIPoolingLayerKernel / @ref CLROIPoolingLayer - @ref CLReductionOperationKernel / @ref CLReductionOperation - CLReshapeLayerKernel / @ref CLReshapeLayer v17.06 Public major release - Various bug fixes - Added support for fixed point 8 bit (QS8) to the various Arm® Neon™ machine learning kernels. - Added unit tests and benchmarks (AlexNet, LeNet) - Added support for sub tensors. - Added infrastructure to provide GPU specific optimisation for some OpenCL kernels. - Added @ref OMPScheduler (OpenMP) scheduler for Neon - Added @ref SingleThreadScheduler scheduler for Arm® Neon™ (For bare metal) - User can specify his own scheduler by implementing the @ref IScheduler interface. - New OpenCL kernels / functions: - @ref CLBatchNormalizationLayerKernel / @ref CLBatchNormalizationLayer - CLDepthConcatenateLayerKernel / CLDepthConcatenateLayer - CLHOGOrientationBinningKernel CLHOGBlockNormalizationKernel, CLHOGDetectorKernel / CLHOGDescriptor CLHOGDetector CLHOGGradient CLHOGMultiDetection - CLLocallyConnectedMatrixMultiplyKernel / CLLocallyConnectedLayer - @ref CLWeightsReshapeKernel / @ref CLConvolutionLayerReshapeWeights - New C++ kernels: - CPPDetectionWindowNonMaximaSuppressionKernel - New Arm® Neon™ kernels / functions: - @ref NEBatchNormalizationLayerKernel / @ref NEBatchNormalizationLayer - NEDepthConcatenateLayerKernel / NEDepthConcatenateLayer - NEDirectConvolutionLayerKernel / @ref NEDirectConvolutionLayer - NELocallyConnectedMatrixMultiplyKernel / NELocallyConnectedLayer - @ref NEWeightsReshapeKernel / @ref NEConvolutionLayerReshapeWeights v17.05 Public bug fixes release - Various bug fixes - Remaining of the functions ported to use accurate padding. - Library does not link against OpenCL anymore (It uses dlopen / dlsym at runtime instead to determine whether or not OpenCL is available). - Added "free" method to allocator. - Minimum version of g++ required for armv7 Linux changed from 4.8 to 4.9 v17.04 Public bug fixes release The following functions have been ported to use the new accurate padding: - CLColorConvertKernel - CLEdgeNonMaxSuppressionKernel - CLEdgeTraceKernel - CLGaussianPyramidHorKernel - CLGaussianPyramidVertKernel - CLGradientKernel - NEChannelCombineKernel - NEFillArrayKernel - NEGaussianPyramidHorKernel - NEGaussianPyramidVertKernel - NEHarrisScoreFP16Kernel - NEHarrisScoreKernel - NEHOGDetectorKernel - NELogits1DMaxKernel - NELogits1DShiftExpSumKernel - NELogits1DNormKernel - NENonMaximaSuppression3x3FP16Kernel - NENonMaximaSuppression3x3Kernel v17.03.1 First Major public release of the sources - Renamed the library to arm_compute - New CPP target introduced for C++ kernels shared between Arm® Neon™ and CL functions. - New padding calculation interface introduced and ported most kernels / functions to use it. - New OpenCL kernels / functions: - CLGEMMLowpMatrixMultiplyKernel / CLGEMMLowp - New Arm® Neon™ kernels / functions: - @ref NENormalizationLayerKernel / @ref NENormalizationLayer - NETransposeKernel / @ref NETranspose - NELogits1DMaxKernel, NELogits1DShiftExpSumKernel, NELogits1DNormKernel / @ref NESoftmaxLayer - @ref NEIm2ColKernel, @ref NECol2ImKernel, NEConvolutionLayerWeightsReshapeKernel / @ref NEConvolutionLayer - NEGEMMMatrixAccumulateBiasesKernel / @ref NEFullyConnectedLayer - @ref NEGEMMLowpMatrixMultiplyKernel / NEGEMMLowp v17.03 Sources preview - New OpenCL kernels / functions: - CLGradientKernel, CLEdgeNonMaxSuppressionKernel, CLEdgeTraceKernel / CLCannyEdge - GEMM refactoring + FP16 support: CLGEMMInterleave4x4Kernel, CLGEMMTranspose1xWKernel, @ref CLGEMMMatrixMultiplyKernel, CLGEMMMatrixAdditionKernel / @ref CLGEMM - CLGEMMMatrixAccumulateBiasesKernel / @ref CLFullyConnectedLayer - CLTransposeKernel / @ref CLTranspose - CLLKTrackerInitKernel, CLLKTrackerStage0Kernel, CLLKTrackerStage1Kernel, CLLKTrackerFinalizeKernel / CLOpticalFlow - @ref CLNormalizationLayerKernel / @ref CLNormalizationLayer - CLLaplacianPyramid, CLLaplacianReconstruct - New Arm® Neon™ kernels / functions: - NEActivationLayerKernel / @ref NEActivationLayer - GEMM refactoring + FP16 support (Requires armv8.2 CPU): @ref NEGEMMInterleave4x4Kernel, @ref NEGEMMTranspose1xWKernel, @ref NEGEMMMatrixMultiplyKernel, @ref NEGEMMMatrixAdditionKernel / @ref NEGEMM - NEPoolingLayerKernel / @ref NEPoolingLayer v17.02.1 Sources preview - New OpenCL kernels / functions: - CLLogits1DMaxKernel, CLLogits1DShiftExpSumKernel, CLLogits1DNormKernel / @ref CLSoftmaxLayer - CLPoolingLayerKernel / @ref CLPoolingLayer - @ref CLIm2ColKernel, @ref CLCol2ImKernel, CLConvolutionLayerWeightsReshapeKernel / CLConvolutionLayer - @ref CLRemapKernel / @ref CLRemap - CLGaussianPyramidHorKernel, CLGaussianPyramidVertKernel / CLGaussianPyramid, CLGaussianPyramidHalf, CLGaussianPyramidOrb - CLMinMaxKernel, CLMinMaxLocationKernel / CLMinMaxLocation - CLNonLinearFilterKernel / CLNonLinearFilter - New Arm® Neon™ FP16 kernels (Requires armv8.2 CPU) - NEAccumulateWeightedFP16Kernel - NEBox3x3FP16Kernel - NENonMaximaSuppression3x3FP16Kernel v17.02 Sources preview - New OpenCL kernels / functions: - CLActivationLayerKernel / @ref CLActivationLayer - CLChannelCombineKernel / CLChannelCombine - CLDerivativeKernel / CLChannelExtract - CLFastCornersKernel / CLFastCorners - CLMeanStdDevKernel / CLMeanStdDev - New Arm® Neon™ kernels / functions: - HOG / SVM: NEHOGOrientationBinningKernel, NEHOGBlockNormalizationKernel, NEHOGDetectorKernel, NEHOGNonMaximaSuppressionKernel / NEHOGDescriptor, NEHOGDetector, NEHOGGradient, NEHOGMultiDetection - NENonLinearFilterKernel / NENonLinearFilter - Introduced a CLScheduler to manage the default context and command queue used by the runtime library and create synchronisation events. - Switched all the kernels / functions to use tensors instead of images. - Updated documentation to include instructions to build the library from sources. v16.12 Binary preview release - Original release */ } // namespace arm_compute