AgeCommit message (Expand)Author
7 hoursFix the OpenCL convolution layer heuristic for int8HEADmasterGian Marco Iodice
12 hoursRename NEGEMMAssembly to CpuGemmAssemblySang-Hoon Park
23 hoursFix Macros.h include in NEScaleMichele Di Giorgio
26 hoursFix integer overflow and null dereferenceMichalis Spyrou
27 hoursRemove padding from CLChannelShuffleLayerKernelManuel Bottini
27 hoursFix Pooling Layer Bug when input is 1xN sizeFreddie Liardet
2 daysFix GEMMLowp output stage validation crash when input's first dimension == 1Giorgio Arena
2 daysFix bug in Select operator when input is 1xNFreddie Liardet
2 daysFix 'ARM_DOT_K0XN0' macro redefinedGiorgio Arena
2 daysSuppress unused variable warning that is present on empty backend buildsGeorgios Pinitas
2 daysNEON/RoiAlign/Quantized/QASYMM8_SIGNED mismatchesManuel Bottini
2 daysRemove unused CLCoreRuntimeContextGeorgios Pinitas
3 daysAdding S32 Support to NEG operator in CLElementwiseUnaryLayerSuhail Munshi
3 daysAdd support for additional CPU variantsGeorgios Pinitas
3 daysAdd constant_weights boolean to FullyConntectedLayerInfoMichele Di Giorgio
3 daysRemove add padding test on weights fot CL DirectConvolutionLayerManuel Bottini
3 daysAdd errata for the performance regressions on OpenCLGian Marco Iodice
7 daysFix missing DATA_TYPE in DOT_PRODUCT4_INTEGER8 OpenCL macroGian Marco Iodice
7 daysRemove TODOsSheri Zhang
7 daysUpdate heuristic for CLConvolutionLayerGian Marco Iodice
7 daysUpdate release noteSheri Zhang
8 daysAdd minimum toolchains requirements to the documentationMichele Di Giorgio
8 daysFix performance issues in NEReductionPablo Marquez Tello
8 daysUpdate SONAME_VERSION in SConscriptSheri Zhang
8 daysRemove deprecated functionsSheri Zhang
9 daysCreates ClKerneLibrary to serve just as a kernel container without build logicGeorgios Pinitas
9 daysRename Quantization/Dequantization kernels/operators to imperative moodGeorgios Pinitas
9 daysAdding S32 support to CLPixelWiseMultiplicationSuhail Munshi
9 daysFix for tanh at small argument valuesAleksandr Nikolaev
9 daysRestructure Documentation (Part 2)Sang-Hoon Park
9 daysRestructure documentationSheri Zhang
9 daysUpdate operator list part3Sheri Zhang
10 daysAdd Queue to API documentationSang-Hoon Park
10 daysFix bug on CLReductionOperationGiorgio Arena
10 daysTurn EXPECT into ASSERT when testing invalid conditionsMichele Di Giorgio
10 daysNEReduceMean failed on v8.2 debug build for AndroidManuel Bottini
10 daysRename PixelwiseMultiplications to Mul for simplicityGeorgios Pinitas
14 daysAdd support for new CPU variantsGeorgios Pinitas
14 daysUpdate operator list documentation. Part 2.Teresa Charlin
2021-04-30Add optimization for global pooling in pooling_layer.clGian Marco Iodice
2021-04-30[Nightly #1144] NEON/DepthwiseConvolutionLayer/Float/F16/Optimized mismatch o...Manuel Bottini
2021-04-29Remove Global pooling optimizationMichalis Spyrou
2021-04-29Remove OpenCL padding: CLReductionOperationKernelGiorgio Arena
2021-04-29Add config_id for CLPixelWiseMultiplicationGian Marco Iodice
2021-04-29Remove stale/solved TODOsMichele Di Giorgio
2021-04-29Fix Global Pooling failuresMichalis Spyrou
2021-04-28Add Queue supportGeorgios Pinitas
2021-04-28Fix for performance regression on G71 (mate9) for direct convolutionAleksandr Nikolaev
2021-04-28Properly parse clvk option in validate examplesMichalis Spyrou
2021-04-28Add per-channel quantization support for CLDeconvolutionLayerFreddie Liardet