Age | Commit message (Expand) | Author |
2018-11-02 | COMPMID-802 Fix NEIm2Col NHWC | Giorgio Arena |
2018-11-02 | COMPMID-959 Refactor OpenCL interceptors to use lambda functions | Anthony Barbier |
2018-11-02 | COMPMID-1012 : Add OpenCL benchmarks for Winograd | Vidhya Sudhan Loganathan |
2018-11-02 | COMPMID-855: Get the library to work on non Mali GPUs | Anthony Barbier |
2018-11-02 | COMPMID-1026 - Add support for 4x4 output tile in CLWinogradConvolutionLayer | Gian Marco Iodice |
2018-11-02 | COMPMID-1037 Add support for F(4x4, 5x5) in CLWinogradOutputTransformKernel | Giorgio Arena |
2018-11-02 | COMPMID-797 Integrate Mobilenet QASYMM8 with new graph. | Giorgio Arena |
2018-11-02 | COMPMID-987: Make beta and gamma optional in BatchNormalization | Michele Di Giorgio |
2018-11-02 | COMPMID-577: Implement CL validation for GaussianPyramid | Sanghoon Lee |
2018-11-02 | COMPMID-1056 - Optimizing CLGEMMMatrixMultiplyKernel refactoring the inner loop | Gian Marco Iodice |
2018-11-02 | COMPMID-761: Add CL/NEON Histogram benchmark tests | John Richardson |
2018-11-02 | COMPMID-959: Add accessors for the OpenCL program cache | Anthony Barbier |
2018-11-02 | COMPMID-922 - CLGEMM FP16 optimizations - part1 | Gian Marco Iodice |
2018-11-02 | COMPMID-811 Add NHWC data format support for CL depthwise convolution QASYMM8 | Giorgio Arena |
2018-11-02 | COMPMID-1037 Add support for F(4x4, 5x5) in CLWinogradInputTransformKernel | Giorgio Arena |
2018-11-02 | COMPMID-959 - Fixing subtensor validation | Gian Marco Iodice |
2018-11-02 | COMPMID-1042: Graph / Stream not working on x86 | Georgios Pinitas |
2018-11-02 | COMPMID-959: Fixed order of init/destruction of CLSymbols / CLKernelLibrary | Anthony Barbier |
2018-11-02 | COMPMID-1031: Use LWS hints for G51, G51BIG, G51LIT, and TNOX | Sam Laynton |
2018-11-02 | COMPMID-761: Add CL/NEON EqualizeHistogram benchmark tests | John Richardson |
2018-11-02 | COMPMID-1054 Update RSH's GEMM to add batch+multi support | Michalis Spyrou |
2018-11-02 | COMPMID-812 Add NHWC data format support for NEON depthwise convolution (opti... | Giorgio Arena |
2018-11-02 | COMPMID-1037 Add support for F(4x4, 5x5) in CLWinogradFilterTransformKernel | Giorgio Arena |
2018-11-02 | COMPMID-959: Add FP32 support to GLES GEMMConvolution | Michele Di Giorgio |
2018-11-02 | COMPMID-1051 - Fix validate method in NEGEMMConvolutionLayer | Gian Marco Iodice |
2018-11-02 | COMPMID-959: Disabled winograd on Arm Cortex-A53 while we're working on a fix | Anthony Barbier |
2018-11-02 | COMPMID-959: Fixed clang-tidy formatting, made GLES builds fail for standalone=1 | Anthony Barbier |
2018-11-02 | COMPMID-584: Add validation to channel_combine kernels | Ioan-Cristian Szabo |
2018-11-02 | COMPMID-802 Add NHWC data format support for NEON im2col. | Giorgio Arena |
2018-11-02 | COMPMID-585: Port OpticalFlow to new validation | John Richardson |
2018-11-02 | COMPMID-1045 - Fix shape_to_valid_region_gaussian_pyramid_half() | Gian Marco Iodice |
2018-11-02 | COMPMID-959: Number of threads was overwritten in benchmark_graph tests | Anthony Barbier |
2018-11-02 | COMPMID-959 Removed assert on cpuid because valgrind always passes '-1' | Anthony Barbier |
2018-11-02 | COMPMID-959 - Fix doxygem comment in CLGEMMConvolutionLayer | Isabella Gottardi |
2018-11-02 | COMPMID-1013 - Create WinogradInfo data structure | Gian Marco Iodice |
2018-11-02 | COMPMID-994 : Check cl_arm_printf is supported in the CLScheduler | Vidhya Sudhan Loganathan |
2018-11-02 | COMPMID-949: Optimizing CLDepthwiseConvolution3x3Kernel for FP16 | Michele Di Giorgio |
2018-11-02 | COMPMID-1040: Added support for nullptr bias tensor in NEWinogradLayer | Andrew Mundy |
2018-11-02 | COMPMID-908 - Merge Activation layer with Convolution Layer (NEON. CL, GLES) | Isabella Gottardi |
2018-11-02 | COMPMID-797: Switch to new graph. | Georgios Pinitas |
2018-11-02 | COMPMID-959 Change DeConvolution weights in doxygen comments | Michalis Spyrou |
2018-11-02 | COMPMID-734: CLTuner rework | Georgios Pinitas |
2018-11-02 | COMPMID-1021: CPUInfo refactoring. | Pablo Tello |
2018-11-02 | COMPMID-998: Release unused trainable parameters. | Georgios Pinitas |
2018-11-02 | COMPMID-596: Port HOGDetector to new validation | John Richardson |
2018-11-02 | COMPMID-1017: Implement dilated convolution in NEON, OpenCL, and GC | Alex Gilday |
2018-11-02 | COMPMID-959: Collapse window in activation layer for GLES. | Georgios Pinitas |
2018-11-02 | COMPMID-1032 - Fixing bug in CLGEMM when is_interleaved_transposed=true | Gian Marco Iodice |
2018-11-02 | COMPMID-959 Fix clang-tidy issue | Anthony Barbier |
2018-11-02 | COMPMID-1016: Optimize kernel reconfiguration | Georgios Pinitas |