| COMPMID-765 - Used GEMM-based convolution in VGG16 |
|
More...
|
over 8 years ago
|
| COMPMID-901 - Optimizing CLCol2ImKernel |
|
More...
|
over 8 years ago
|
| COMPMID-879: Investigate CL mismatches in Convolution S16 |
|
More...
|
over 8 years ago
|
| COMPMID-897 Merge batch normalization with bounded relu |
|
More...
|
over 8 years ago
|
| COMPMID-893: Fix segfault in graph_squeezenet_v1_1 |
|
More...
|
over 8 years ago
|
| COMPMID-765 - Added LWS hint in CLIm2Col |
|
More...
|
over 8 years ago
|
| COMPMID-896: Replace legacy 4x4 u8 GEMM kernel with safe version. |
|
More...
|
almost 9 years ago
|
| COMPMID-887 Valgrind invalid read in print_measurements |
|
More...
|
almost 9 years ago
|
| COMPMID-845: Create a ConvolutionLayer for CL |
|
More...
|
almost 9 years ago
|
| COMPMID-876: Integrate RSH native GEMM kernel. |
|
More...
|
almost 9 years ago
|
| COMPMID-894: Segfault: neon_cnn on S5 neo |
|
More...
|
almost 9 years ago
|
| COMPMID-784: Added support for PADDING = SAME in Winograd layer. |
|
More...
|
almost 9 years ago
|
| COMPMID-765: Add Winograd Convolution Hint for NEON Convolution |
|
More...
|
almost 9 years ago
|
| COMPMID-892: OCLGrind failures on both validation and benchmark |
|
More...
|
almost 9 years ago
|
| COMPMID-884: Valgrind: NEDirectConvolutionLayerKernel invalid read |
|
More...
|
almost 9 years ago
|
| COMPMID-888 Valgrind invalid read in NEGEMVAArch64Kernel |
|
More...
|
almost 9 years ago
|
| COMPMID-597: Port HOGMultiDetection to new framework |
|
More...
|
almost 9 years ago
|
| COMPMID-765 - Extended GEMM benchmark |
|
More...
|
almost 9 years ago
|
| COMPMID-885 Valgrind NEConvolutionSquare::configure relies on uninitialised values |
|
More...
|
almost 9 years ago
|
| COMPMID-765 - Used GEMM instead of Direct into the GoogleNet graph example |
|
More...
|
almost 9 years ago
|
| COMPMID-846: Create a ConvolutionLayer for NEON |
|
More...
|
almost 9 years ago
|
| COMPMID-754: Add validation method to CLPermute kernel |
|
More...
|
almost 9 years ago
|
| COMPMID-792: Update doctrings of CL kernels supporting broadcast operations |
|
More...
|
almost 9 years ago
|
| COMPMID-856: CL Depthwise Convolution QASYMM8 support |
|
More...
|
almost 9 years ago
|
| COMPMID-890: Valgrind: NEON Convolution Layer validation fails |
|
More...
|
almost 9 years ago
|
| COMPMID-889 - Implement Squeezenet 1.1 as a graph example |
|
More...
|
almost 9 years ago
|
| COMPMID-765: Fixed clangtidy warnings. |
|
More...
|
almost 9 years ago
|
| COMPMID-874: Improve default number of threads choice in the Scheduler |
|
More...
|
almost 9 years ago
|
| COMPMID-895 - Optimizing CLDepthwiseConvolution3x3Kernel |
|
More...
|
almost 9 years ago
|
| COMPMID-891 - Use OpenCL timer in CLTuner |
|
More...
|
almost 9 years ago
|