opencv

Commit Graph

Author	SHA1	Message	Date
Alexander Alekhin	eab6744ac7	dnn(ocl): use compile-time LOCAL_SIZE parameter instead of get_local_size(0) and dynamic local memory allocation	6 years ago
Dmitry Kurtaev	c918ac298c	Fix IE tests	6 years ago
Dmitry Kurtaev	ac262f5b5d	Clone convolution layer weights only for fusion	6 years ago
Dmitry Kurtaev	3c3c5ef2b6	Fix a dnn bug with retrieving all the output blobs	6 years ago
Dmitry Kurtaev	ff775b2e54	Remove ASSERT_ANY_THROW checks fpr Myriad plugin and FP32 networks	6 years ago
Alexander Nesterov	97c3bcb1b7	Added fix for other size	6 years ago
Lee Jaehwan	3721c8bb06	Merge pull request #13586 from eightco:Core_bugfix3 * Add Operator override for multi-channel Mat with literal constant. * simple test * Operator overloading channel constraint for primitive types * fix some test for #13586	6 years ago
Dmitry Kurtaev	f0ddf302b2	Move Inference Engine to new API	6 years ago
Raphael Graf	82c77fa244	dnn: remove malloc.h include	6 years ago
Dmitry Kurtaev	d0504c95f4	Add a text message for Convolution layer's input channels check	6 years ago
WuZhiwen	3d44e9ad92	Merge pull request #13520 from wzw-intel:hang * dnn/Vulkan: fix GPU hang for heavy convolution tasks Intel i915 driver will declare GPU hang if the compute shader takes too long to complete. See https://bugs.freedesktop.org/show_bug.cgi?id=108947 for details. The idea in this commit is to divide heavy task into several light ones and run compute shader multiple times to make each run take short time enough. TODO: Add more efficient compute shader Signed-off-by: Wu Zhiwen <zhiwen.wu@intel.com> * dnn/Vulkan: add a more efficient conv shader	6 years ago
Wu Zhiwen	dae03273cd	dnn: fixup missing vkcom/vulkan combination of backend/target in dnn test	6 years ago
Wu Zhiwen	be6a837e15	dnn: add Vulkan device check for BackendRegistry	6 years ago
Dmitry Kurtaev	840c892abd	Batch normalization in training phase from Torch	6 years ago
Dmitry Kurtaev	59ce1d80a5	Fix dnn tests for Inference Engine R5	6 years ago
Dmitry Kurtaev	257f60582a	Add serialize method for IE net wrapper backport `4ba4901ca9`	6 years ago
Alexander Alekhin	bbdc987fc6	dnn: add OpenVINO 2018R5 defines https://software.intel.com/en-us/openvino-toolkit	6 years ago
Dmitry Kurtaev	e71758cfdf	Operate with shapes in ONNX models	6 years ago
Alexander Alekhin	d8583b2c7a	dnn: fix vulkan backend builds with Clang	6 years ago
Dmitry Kurtaev	53f6198f27	Minor fixes in IE backend tests	6 years ago
Dmitry Kurtaev	8422dda2c7	Element-wise subtraction from TensorFlow	6 years ago
Dmitry Kurtaev	93971a53d9	Exclude Input layer from list of outputs for IE networks	6 years ago
Dmitry Kurtaev	3868cb44f1	Exclude Input layer from list of outputs for IE networks	6 years ago
Alexander Alekhin	9ff1c39daa	dnn: fixup available backends/targets	6 years ago
Maksim Shabunin	fe459c82e5	Merge pull request #13332 from mshabunin:dnn-backends DNN backends registry (#13332) * Added dnn backends registry * dnn: process DLIE/FPGA target	6 years ago
Dmitry Kurtaev	c9e0c77d73	Concat layer from TensorFlow with constant inputs	6 years ago
Dmitry Kurtaev	4ba4901ca9	Add serialize method for IE net wrapper	6 years ago
Dmitry Kurtaev	84ce2cc211	Enable some dnn tests according to the new Intel's Inference Engine release (R4)	6 years ago
Wu Zhiwen	4e65283081	dnn/Vulkan: make thread safe Use a global dedicated mutex to make sure initialize once and protect command buffer pool and queue. Signed-off-by: Wu Zhiwen <zhiwen.wu@intel.com>	6 years ago
Dmitry Kurtaev	2f6f52d644	Fix ONNX's emotion_ferplus model. Reduce input size for OpenPose tests	6 years ago
WuZhiwen	02cc1cd6e6	Merge pull request #13244 from wzw-intel:init_vulkan * dnn/Vulkan: don't init Vulkan runtime if using other backend/target Don't need to explictly call a init API but will automatically init Vulkan environment the first time to use an VkCom object. Signed-off-by: Wu Zhiwen <zhiwen.wu@intel.com> * dnn/Vulkan: depress compilier warning for "-Wsign-promo" Signed-off-by: Wu Zhiwen <zhiwen.wu@intel.com>	6 years ago
Alexander Alekhin	d7272f76fb	dnn: fix format	6 years ago
Dmitry Kurtaev	0d117312c9	DNN_TARGET_FPGA using Intel's Inference Engine	6 years ago
Alexander Alekhin	dd3398416b	experimental version++	6 years ago
Christopher Gundler	b58a8729c2	Merge pull request #13131 from Christopher22:add_transposedConv_onnx * Add support for ConvTranspose when parsing ONNX. * Add support for ConvTranspose when parsing ONNX. * Add test for Deconvolution	6 years ago
Dmitry Kurtaev	b5c54e447c	Extra hyperparameters for Intel's Inference Engine layers	6 years ago
Dmitry Kurtaev	ef5d921eac	Fix Vulkan's max pooling in case of no output indices	6 years ago
Alexander Alekhin	96c71dd3d2	dnn: reduce set of ignored warnings	6 years ago
catree	10b482ff1e	Fix code and missing intrin header. Remove useless header.	6 years ago
Dmitry Kurtaev	80265a0815	Fix a bug with OpenVINO backend	6 years ago
Alexander Alekhin	801c943009	fix coverity reports	6 years ago
Dmitry Kurtaev	dc9e6d3af8	Update a script to generate text graphs for Faster-RCNN networks from TensorFlow	6 years ago
catree	eebf0dd7c9	Fix integer overflow when accumulating timing values.	6 years ago
Wu Zhiwen	33c9d57c6f	dnn/Vulkan: skip heavy convolution task This is a workaround for GPU hang on heavy convolution workload (> 10 GFLOPS). e.g. ResNet101_DUC_HDC For the long time task, vkWaitForFences() return without error but next call on vkQueueSubmit() return -4, i.e. "VK_ERROR_DEVICE_LOST" and driver reports GPU hang. Need more investigation on root cause of GPU hang and need to optimize convolution shader to reduce process time.	6 years ago
Dmitry Kurtaev	a6f9170f10	Add ONNX's padding import	6 years ago
Wu Zhiwen	34e9d1eb3c	dnn/Vulkan: support log softmax Signed-off-by: Wu Zhiwen <zhiwen.wu@intel.com>	6 years ago
Wu Zhiwen	3914c17b0d	dnn/Vulkan: Refine error handle mechanism Fallback to OPENCV backend and CPU target if catch exception from vkcom backend. Signed-off-by: Wu Zhiwen <zhiwen.wu@intel.com>	6 years ago
Wu Zhiwen	7fff245f87	dnn/Vulkan: Rename function_list.inl Signed-off-by: Wu Zhiwen <zhiwen.wu@intel.com>	6 years ago
WuZhiwen	6e3ea8b49d	Merge pull request #12703 from wzw-intel:vkcom * dnn: Add a Vulkan based backend This commit adds a new backend "DNN_BACKEND_VKCOM" and a new target "DNN_TARGET_VULKAN". VKCOM means vulkan based computation library. This backend uses Vulkan API and SPIR-V shaders to do the inference computation for layers. The layer types that implemented in DNN_BACKEND_VKCOM include: Conv, Concat, ReLU, LRN, PriorBox, Softmax, MaxPooling, AvePooling, Permute This is just a beginning work for Vulkan in OpenCV DNN, more layer types will be supported and performance tuning is on the way. Signed-off-by: Wu Zhiwen <zhiwen.wu@intel.com> * dnn/vulkan: Add FindVulkan.cmake to detect Vulkan SDK In order to build dnn with Vulkan support, need installing Vulkan SDK and setting environment variable "VULKAN_SDK" and add "-DWITH_VULKAN=ON" to cmake command. You can download Vulkan SDK from: https://vulkan.lunarg.com/sdk/home#linux For how to install, see https://vulkan.lunarg.com/doc/sdk/latest/linux/getting_started.html https://vulkan.lunarg.com/doc/sdk/latest/windows/getting_started.html https://vulkan.lunarg.com/doc/sdk/latest/mac/getting_started.html respectively for linux, windows and mac. To run the vulkan backend, also need installing mesa driver. On Ubuntu, use this command 'sudo apt-get install mesa-vulkan-drivers' To test, use command '$BUILD_DIR/bin/opencv_test_dnn --gtest_filter=VkCom' Signed-off-by: Wu Zhiwen <zhiwen.wu@intel.com> * dnn/Vulkan: dynamically load Vulkan runtime No compile-time dependency on Vulkan library. If Vulkan runtime is unavailable, fallback to CPU path. Use environment "OPENCL_VULKAN_RUNTIME" to specify path to your own vulkan runtime library. Signed-off-by: Wu Zhiwen <zhiwen.wu@intel.com> * dnn/Vulkan: Add a python script to compile GLSL shaders to SPIR-V shaders The SPIR-V shaders are in format of text-based 32-bit hexadecimal numbers, and inserted into .cpp files as unsigned int32 array. * dnn/Vulkan: Put Vulkan headers into 3rdparty directory and some other fixes Vulkan header files are copied from https://github.com/KhronosGroup/Vulkan-Docs/tree/master/include/vulkan to 3rdparty/include Fix the Copyright declaration issue. Refine OpenCVDetectVulkan.cmake * dnn/Vulkan: Add vulkan backend tests into existing ones. Also fixed some test failures. - Don't use bool variable as uniform for shader - Fix dispathed group number beyond max issue - Bypass "group > 1" convolution. This should be support in future. * dnn/Vulkan: Fix multiple initialization in one thread.	6 years ago
Maksim Shabunin	e6d9486a6c	Fixed several issues found by static analysis	6 years ago

... 7 8 9 10 11 ...

1398 Commits (29976a04c5cbcf5a65e55863cd2628a3be99e2e5)