Gitiles
Code Review
Sign In
review.mlplatform.org
/
ml
/
ComputeLibrary
/
503cec81c77eaa0d7f6b98358de287bbfe04bcb5
/
src
33d2207
COMPMID-860: Neon HGEMM integrated assembly kernel from RSH for Arm Cortex-A55r1.
by Pablo Tello
· 7 years ago
e8a4983
COMPMID-798 Add instrumentation to NEON kernels
by Anthony Barbier
· 7 years ago
7b4d547
COMPMID-816 - Optimizing CLGEMMLowpMatrixMultiplyCore - Part1
by Gian Marco
· 7 years ago
dde3ad9
COMPMID-842: Add NEON QASYMM8 RELU Activation
by Michele Di Giorgio
· 7 years ago
967f86d
COMPMID-858: Assert in ICLKernel on higher window dimensions moved to enqueue
by Anthony Barbier
· 7 years ago
a1f7e33
COMPMID-841: Add CL QASYMM8 RELU Activation
by Michele Di Giorgio
· 7 years ago
6c6e77a
COMPMID-784: Doxygen fixes
by Pablo Tello
· 7 years ago
e855c23
APPBROWSER-377: GCConvoutionLayer support for FP16
by Stephen Li
· 7 years ago
9ceebbe
COMPMID-815: Updated NEWinogradLayer with the lastest code from Research.
by Pablo Tello
· 7 years ago
00afd11
COMPMID-719: NEPermuteKernel refactoring
by Pablo Tello
· 7 years ago
5237e01
COMPMID-838 Implement CLPermute
by Michalis Spyrou
· 7 years ago
652bde5
COMPMID-674 - Create Google InceptionV3 example
by Georgios Pinitas
· 7 years ago
f72f936
COMPMID-791: Adds support of QASYMM8 in NEDepthwiseConvolution3x3
by Georgios Pinitas
· 7 years ago
61ef5bf
IVGCVSW-847 Fix {NEON/CL}PoolingLayerKernel config
by Diego Lopez Recas
· 7 years ago
a0d1183
COMPMID-751 QASYMM8 ActivationLayer optimisation: don't requantize if not necessary
by Giorgio Arena
· 7 years ago
944d3f7
COMPMID-751 Processing 8 elements makes computation up to 80us faster on MobileNet QASYMM8 dwc layers
by Giorgio Arena
· 7 years ago
a66eaa2
COMPMID-752 Creating an example for QASYMM8 MobileNet
by Giorgio Arena
· 7 years ago
621965e
COMPMID-769 Add asymmetric padding support in NEON kernels.
by Michalis Spyrou
· 7 years ago
20d7848
COMPMID-816 - Enabled CLConvolutionLayer to use CLGEMM function instead
by Gian Marco
· 7 years ago
84f3ae8
COMPMID-830 Fix hang in arm_compute_benchmark NEON
by Michalis Spyrou
· 7 years ago
5518671
COMPMID-785: Add QASYMM8 support for pooling layer
by Georgios Pinitas
· 7 years ago
1ff480b
COMPMID-765: Fixed unused variable warning
by Anthony Barbier
· 7 years ago
093adbc
COMPMID-753 Add benchmarks for GEMM/GEMMLowp used in AlexNet
by Giorgio Arena
· 7 years ago
bcbc970
IVGCVSW-863 calculate_max_window..() family takes ValidRegion
by Diego Lopez Recas
· 7 years ago
1bfc784
APPBROWSER-376: Work around for scale validation error.
by Frank Lei
· 7 years ago
780db4e
COMPMID-471 Implement Deconvolution on OpenCL
by Michalis Spyrou
· 7 years ago
d7ba539
COMPMID-727 - Implement reference and CL/NEON validation for CustomConvolutionRectangle
by Sanghoon Lee
· 7 years ago
fde97fb
COMPMID-784: Removed no longer needed file winograd_shim_nchw.hpp
by Pablo Tello
· 7 years ago
4e2c139
COMPMID-784: Winograd refactoring
by Pablo Tello
· 7 years ago
fd14f9e
COMPMID-783: Segfault in OCLGrind
by Georgios Pinitas
· 7 years ago
abd03cf
APPBROWSER-298: Remove the old shader common code
by Joel Liang
· 7 years ago
25f2368
COMPMID-589: Port HOGDescriptor to new validation
by John Richardson
· 7 years ago
1d25ed5
COMPMID-759 - CLGEMM optimization for McVail benchmarks
by Gian Marco
· 7 years ago
57b2010
APPBROWSER-373: Rewrite the convolution_layer.cs with the new common code
by zhenglin
· 7 years ago
6fc06cc
APPBROWSER-374: Rewrite the dropout.cs with the common code
by zhenglin
· 7 years ago
b91e34c
COMPMID-746 Allow NEDirectConvolution to work without biases for QS.
by Michalis Spyrou
· 7 years ago
aa1209a
APPBROWSER-375: Rewrite the transpose.cs with the new common code
by zhenglin
· 7 years ago
09849a0
APPBROWSER-372: Rewrite the direct_convolution5x5.cs with the new common code
by Joel Liang
· 7 years ago
07d4054
APPBROWSER-370: Rewrite the normalization_layer.cs with the new common code
by zhenglin
· 7 years ago
6485691
COMPMID-773: Add CL/NEON Harris Corners benchmark tests
by Alex Gilday
· 7 years ago
384a43e
COMPMID-579: Port Derivative to new validation
by John Richardson
· 7 years ago
6387543
APPBROWSER-371: Rewrite the direct_convolution3x3.cs with the new common code
by Joel Liang
· 7 years ago
19e9142
APPBROWSER-369: Rewrite the gemm.cs with the new common code
by zhenglin
· 7 years ago
fef6dae
COMPMID-750: Enabled support for U8 and S8 datatypes in NEGEMMLowpAArch64V8P4Kernel
by Pablo Tello
· 7 years ago
8cdfdb8
APPBROWSER-366: Add DepthwiseConvolutionLayer(fp16 only) support.
by Frank Lei
· 7 years ago
02541fb
COMPMID-719: NEWinogradLayer reordering using NEPermute.
by Pablo Tello
· 7 years ago
d064389
APPBROWSER-357: Fix Transpose performance issue by tuning lws
by steli01
· 7 years ago
1d08a31
COMPMID-765: Collapse execution window in CL kernels.
by Georgios Pinitas
· 7 years ago
1250a5a
COMPMID-767 : Propagate hints to subgraph.
by Georgios Pinitas
· 7 years ago
7c23ad0
COMPMID-617 Add validation to NEConvolutionLayer
by Giorgio Arena
· 7 years ago
926d5e1
APPBROWSER-356: Implement GCArithmeticAddition Operator
by zhenglin
· 7 years ago
358ca20
COMPMID-617: Adds CLFullyConnectionLayer validation support
by Georgios Pinitas
· 7 years ago
eb8f71e
APPBROWSER-230, APPBROWSER-300: Rename GCScheduler enqueue to dispatch, sync to memory_barrier
by Joel Liang
· 7 years ago
1c5ffd6
APPBROWSER-359: Sync tensor in GC benchmark tests
by Joel Liang
· 7 years ago
c5a7e59
APPBROWSER-365: Rewrite the pooling_layer.cs with the new common code
by Joel Liang
· 7 years ago
08d5421
APPBROWSER-363: Rewrite the ActivationLayer.cs with the new common code
by zhenglin
· 7 years ago
82588e8
APPBROWSER-364: Rewrite the DepthConcatLayer.cs with the new common code
by zhenglin
· 7 years ago
57a150a
APPBROWSER-341: Add ResizeNearest operator support
by Frank Lei
· 7 years ago
2f41981
APPBROWSER-358: Fix shader compilation error in matrix addition
by Frank Lei
· 7 years ago
a9e1533
COMPMID-766 Allow graph examples to run without OpenCL being present on the platform
by Anthony Barbier
· 7 years ago
77eb21f
Fix destruction order of singleton objects and resolve hang GLES issue.
by Ioan-Cristian Szabo
· 7 years ago
490b3d8
IVGCVSW-863 Move clamp, for_each, foldl to misc/utility.h
by Diego Lopez Recas
· 7 years ago
53a6ec5
APPBROWSER-304,342: Add exclude padding support for OpenGL ES implementation and implement MaxPool operators
by Xinghang Zhou
· 7 years ago
aa62b28
APPBROWSER-344: Change data type of the SSBOs to improve the Softmax performance
by Joel Liang
· 7 years ago
0fb6cf5
APPBROWSER-340: Implement NormalizePlanarYUV operator
by zhenglin
· 7 years ago
08c5a06
COMPMID-750: Fix assembly kernel interfaces
by Georgios Pinitas
· 7 years ago
7b43d31
COMPMID-556: Don't call dlclose
by Anthony Barbier
· 7 years ago
cc9fed5
COMPMID-747 Fixed AccessWindow in CL Direct convolution
by Anthony Barbier
· 7 years ago
b660dcf
COMPMID-556: Fix NEConvolutionLayer for fully connected convolution.
by Georgios Pinitas
· 7 years ago
5712091
Add support to print tensor with multiple channels.
by Ioan-Cristian Szabo
· 7 years ago
b5311a6
COMPMID-556 - Enabled optimized path for Bifrost in CLGEMM
by Gian Marco
· 7 years ago
0223a78
COMPMID-556: Fix bugs around NEDirectConvolutionLayer
by Georgios Pinitas
· 7 years ago
1568621
COMPMID-744 Fixed access windows in DepthwiseConvolutionLayer
by Anthony Barbier
· 7 years ago
93b9bdb
COMPMID-743 Fixed AccessWindow in GEMMLowpMatrixMultiply
by Anthony Barbier
· 7 years ago
4062680
COMPMID-556: Fix checks in NEGEMMLowpOffsetContributionKernel
by Georgios Pinitas
· 7 years ago
9d36484
COMPMID-556: Force clone() of SubTensorInfo come from a TensorInfo
by Georgios Pinitas
· 7 years ago
95ca9a5
COMPMID-556 Fixed clang-tidy error
by Anthony Barbier
· 7 years ago
bfa3b52
COMPMID-556 - Fix examples
by Gian Marco
· 7 years ago
c8a85ba
COMPMID-563 - Implement reference and CL/NEON validation for Custom_Convolution (Output U8)
by Sanghoon Lee
· 7 years ago
284c204
COMPMID-556: Allocate trainable data tensor after configuration
by Georgios Pinitas
· 7 years ago
ff85093
COMPMID-741 - Changelog for major release 17.12
by Gian Marco
· 7 years ago
c0ee461
COMPMID-556: Fixed clang tidy warnings
by Anthony Barbier
· 7 years ago
c2708ec
COMPMID-556: Switched from int32 to int to avoid compiling issues with bare_metal
by Anthony Barbier
· 7 years ago
bd5900f
COMPMID-556 Fixed validate for Dot product MatrixMultiplyCore
by Anthony Barbier
· 7 years ago
4708e02
COMPMID-556 - Fixed CLDirectConvolutionLayer mismatches
by Michalis Spyrou
· 7 years ago
5dfeae6
IVGCVSW-820: Add QASYMM8 support to NeonActivationLayerKernel
by Michel Iwaniec
· 7 years ago
7f0f790
COMPMID-731 - Remove padding requirements for NEGEMMLowpOutputStage
by Gian Marco
· 7 years ago
f2ad401
COMPMID-718: Fixed clang warnings
by Pablo Tello
· 7 years ago
7d473dd
APPBROWSER-323: Transpose performance optimization
by steli01
· 7 years ago
b42d53c
COMPMID-732 - Remove padding requirements in NETranspose
by Gian Marco
· 7 years ago
afa5d81
COMPMID-617 Add validation methods to Kernels
by Michalis Spyrou
· 7 years ago
631c41a
COMPMID-556: Rename Error to Status and inverse logic
by Georgios Pinitas
· 7 years ago
57f249b
COMPMID-661: Add Bifrost lws heuristics dwc with pad=1 [new_dev] #52
by Georgios Pinitas
· 7 years ago
a2b89ca
IVGCVSW-631 Neon support for Softmax beta parameter (F32 only)
by Pablo Palmier
· 7 years ago
5948634
COMPMID-617 Add window validation to CLDirectConvolutionLayer
by Giorgio Arena
· 7 years ago
424eb5d
COMPMID-556: Fix dangling references in NEFillBorder
by Georgios Pinitas
· 7 years ago
4af62a0
APPBROWSER-289 DirectConvolution1*1 optimization for FP16, and fix a typo in PoolingLayer
by Xinghang Zhou
· 7 years ago
7c435f2
COMPMID-728 - Added validation for transpose
by Gian Marco
· 7 years ago
07c37f9
APPBROWSER-313: Performance improvement for softmax layer
by Joel Liang
· 7 years ago
5124be5
COMPMID-661: Convolution quantized (#32)
by Chunosov
· 7 years ago
Next »