The ocl convolution doesn't support tensorflow padMode well.
Add fallback check if we meet this situation, it could fix the
tensorflow MobileNet SSD failure.
Signed-off-by: Li Peng <peng.li@intel.com>
* fix faster_rcnn sample crashed at PoolingInvoker operator() of pooling_layer.
* find_odj onmouse bug about find matched point status.
* reverted AutoBuffer back to std::vector
* Added the imagesFromBlob method to the dnn module.
* Rewritten imagesFromBlob based on first dkurt comments
* Updated code with getPlane()
* Modify comment of imagesFromBlob() in dnn module
* modified comments, removed useless assertions & added OutputArrayOfArray
* replaced tabs with whitespaces & put vectorOfChannels instantiation outside the loop
* Changed pre-commit.sample to pre-commit in .git/hooks/
* Added a test for imagesFromBlob in test_misc.cpp (dnn)
* Changed nbOfImages, robustified test with cv::randu, modified assertion
- removed tr1 usage (dropped in C++17)
- moved includes of vector/map/iostream/limits into ts.hpp
- require opencv_test + anonymous namespace (added compile check)
- fixed norm() usage (must be from cvtest::norm for checks) and other conflict functions
- added missing license headers
* Do not build protobuf if dnn is disabled
* Added BUILD_LIST cmake option to the cache
* Moved protobuf to the top level
* Fixed static build
* Fixed world build
* fixup! Fixed world build
* Add a 512 bit codepath to the AVX512 fastConv function
this patch adds a 512 wide codepath to the fastConv() function for
AVX512 use.
The basic idea is to process the first N * 16 elements of the vector
with avx512, and then run the rest of the vector using the traditional
AVX2 codepath.
* dnn: use unaligned AVX512 load (OpenCV aligns data on 32-byte boundary)
* dnn: change "vecsize" condition for AVX512
* dnn: fix indentation
This patch adds AVX512 optimized fastConv as well as the hookups
needed to get these called in the convolution_layer.
AVX512 fastConv is code-identical on a C level to the AVX2 one,
but is measurably faster due to AVX512 having more registers available
to cache results in.
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Incorrect type of negative_slope result in this bug.
Also and OCL test for darknet to validate this patch.
Signed-off-by: Wu Zhiwen <zhiwen.wu@intel.com>
convolution kernel use default queue to run, so that ocl::Timer
, to measure the kernel run time, should use the default queue too.
Also remove useless parameter for convolve()
Signed-off-by: Wu Zhiwen <zhiwen.wu@intel.com>
Add layer forward interface with InputArrayOfArrays and
OutputArrayOfArrays parameters, it allows UMat buffer to be
processed and transferred in the layers.
Signed-off-by: Li Peng <peng.li@intel.com>
These two typdefs are not compiled when BUILD_opencv_dnn is set to
false, however there are other modules that uses these typedef so
it may cause build errors. Moving typedef to the python module
ensures they are always defined.
add libdnn acceleration to dnn module (#9114)
* import libdnn code
Signed-off-by: Li Peng <peng.li@intel.com>
* add convolution layer ocl acceleration
Signed-off-by: Li Peng <peng.li@intel.com>
* add pooling layer ocl acceleration
Signed-off-by: Li Peng <peng.li@intel.com>
* add softmax layer ocl acceleration
Signed-off-by: Li Peng <peng.li@intel.com>
* add lrn layer ocl acceleration
Signed-off-by: Li Peng <peng.li@intel.com>
* add innerproduct layer ocl acceleration
Signed-off-by: Li Peng <peng.li@intel.com>
* add HAVE_OPENCL macro
Signed-off-by: Li Peng <peng.li@intel.com>
* fix for convolution ocl
Signed-off-by: Li Peng <peng.li@intel.com>
* enable getUMat() for multi-dimension Mat
Signed-off-by: Li Peng <peng.li@intel.com>
* use getUMat for ocl acceleration
Signed-off-by: Li Peng <peng.li@intel.com>
* use CV_OCL_RUN macro
Signed-off-by: Li Peng <peng.li@intel.com>
* set OPENCL target when it is available
and disable fuseLayer for OCL target for the time being
Signed-off-by: Li Peng <peng.li@intel.com>
* fix innerproduct accuracy test
Signed-off-by: Li Peng <peng.li@intel.com>
* remove trailing space
Signed-off-by: Li Peng <peng.li@intel.com>
* Fixed tensorflow demo bug.
Root cause is that tensorflow has different algorithm with libdnn
to calculate convolution output dimension.
libdnn don't calculate output dimension anymore and just use one
passed in by config.
* split gemm ocl file
split it into gemm_buffer.cl and gemm_image.cl
Signed-off-by: Li Peng <peng.li@intel.com>
* Fix compile failure
Signed-off-by: Li Peng <peng.li@intel.com>
* check env flag for auto tuning
Signed-off-by: Li Peng <peng.li@intel.com>
* switch to new ocl kernels for softmax layer
Signed-off-by: Li Peng <peng.li@intel.com>
* update softmax layer
on some platform subgroup extension may not work well,
fallback to non subgroup ocl acceleration.
Signed-off-by: Li Peng <peng.li@intel.com>
* fallback to cpu path for fc layer with multi output
Signed-off-by: Li Peng <peng.li@intel.com>
* update output message
Signed-off-by: Li Peng <peng.li@intel.com>
* update fully connected layer
fallback to gemm API if libdnn return false
Signed-off-by: Li Peng <peng.li@intel.com>
* Add ReLU OCL implementation
* disable layer fusion for now
Signed-off-by: Li Peng <peng.li@intel.com>
* Add OCL implementation for concat layer
Signed-off-by: Wu Zhiwen <zhiwen.wu@intel.com>
* libdnn: update license and copyrights
Also refine libdnn coding style
Signed-off-by: Wu Zhiwen <zhiwen.wu@intel.com>
Signed-off-by: Li Peng <peng.li@intel.com>
* DNN: Don't link OpenCL library explicitly
* DNN: Make default preferableTarget to DNN_TARGET_CPU
User should set it to DNN_TARGET_OPENCL explicitly if want to
use OpenCL acceleration.
Also don't fusion when using DNN_TARGET_OPENCL
* DNN: refine coding style
* Add getOpenCLErrorString
* DNN: Use int32_t/uint32_t instread of alias
* Use namespace ocl4dnn to include libdnn things
* remove extra copyTo in softmax ocl path
Signed-off-by: Li Peng <peng.li@intel.com>
* update ReLU layer ocl path
Signed-off-by: Li Peng <peng.li@intel.com>
* Add prefer target property for layer class
It is used to indicate the target for layer forwarding,
either the default CPU target or OCL target.
Signed-off-by: Li Peng <peng.li@intel.com>
* Add cl_event based timer for cv::ocl
* Rename libdnn to ocl4dnn
Signed-off-by: Li Peng <peng.li@intel.com>
Signed-off-by: wzw <zhiwen.wu@intel.com>
* use UMat for ocl4dnn internal buffer
Remove allocateMemory which use clCreateBuffer directly
Signed-off-by: Li Peng <peng.li@intel.com>
Signed-off-by: wzw <zhiwen.wu@intel.com>
* enable buffer gemm in ocl4dnn innerproduct
Signed-off-by: Li Peng <peng.li@intel.com>
* replace int_tp globally for ocl4dnn kernels.
Signed-off-by: wzw <zhiwen.wu@intel.com>
Signed-off-by: Li Peng <peng.li@intel.com>
* create UMat for layer params
Signed-off-by: Li Peng <peng.li@intel.com>
* update sign ocl kernel
Signed-off-by: Li Peng <peng.li@intel.com>
* update image based gemm of inner product layer
Signed-off-by: Li Peng <peng.li@intel.com>
* remove buffer gemm of inner product layer
call cv::gemm API instead
Signed-off-by: Li Peng <peng.li@intel.com>
* change ocl4dnn forward parameter to UMat
Signed-off-by: Li Peng <peng.li@intel.com>
* Refine auto-tuning mechanism.
- Use OPENCV_OCL4DNN_KERNEL_CONFIG_PATH to set cache directory
for fine-tuned kernel configuration.
e.g. export OPENCV_OCL4DNN_KERNEL_CONFIG_PATH=/home/tmp,
the cache directory will be /home/tmp/spatialkernels/ on Linux.
- Define environment OPENCV_OCL4DNN_ENABLE_AUTO_TUNING to enable
auto-tuning.
- OPENCV_OPENCL_ENABLE_PROFILING is only used to enable profiling
for OpenCL command queue. This fix basic kernel get wrong running
time, i.e. 0ms.
- If creating cache directory failed, disable auto-tuning.
* Detect and create cache dir on windows
Signed-off-by: Li Peng <peng.li@intel.com>
* Refine gemm like convolution kernel.
Signed-off-by: Li Peng <peng.li@intel.com>
* Fix redundant swizzleWeights calling when use cached kernel config.
* Fix "out of resource" bug when auto-tuning too many kernels.
* replace cl_mem with UMat in ocl4dnnConvSpatial class
* OCL4DNN: reduce the tuning kernel candidate.
This patch could reduce 75% of the tuning candidates with less
than 2% performance impact for the final result.
Signed-off-by: Zhigang Gong <zhigang.gong@intel.com>
* replace cl_mem with umat in ocl4dnn convolution
Signed-off-by: Li Peng <peng.li@intel.com>
* remove weight_image_ of ocl4dnn inner product
Actually it is unused in the computation
Signed-off-by: Li Peng <peng.li@intel.com>
* Various fixes for ocl4dnn
1. OCL_PERFORMANCE_CHECK(ocl::Device::getDefault().isIntel())
2. Ptr<OCL4DNNInnerProduct<float> > innerProductOp
3. Code comments cleanup
4. ignore check on OCL cpu device
Signed-off-by: Li Peng <peng.li@intel.com>
* add build option for log softmax
Signed-off-by: Li Peng <peng.li@intel.com>
* remove unused ocl kernels in ocl4dnn
Signed-off-by: Li Peng <peng.li@intel.com>
* replace ocl4dnnSet with opencv setTo
Signed-off-by: Li Peng <peng.li@intel.com>
* replace ALIGN with cv::alignSize
Signed-off-by: Li Peng <peng.li@intel.com>
* check kernel build options
Signed-off-by: Li Peng <peng.li@intel.com>
* Handle program compilation fail properly.
* Use std::numeric_limits<float>::infinity() for large float number
* check ocl4dnn kernel compilation result
Signed-off-by: Li Peng <peng.li@intel.com>
* remove unused ctx_id
Signed-off-by: Li Peng <peng.li@intel.com>
* change clEnqueueNDRangeKernel to kernel.run()
Signed-off-by: Li Peng <peng.li@intel.com>
* change cl_mem to UMat in image based gemm
Signed-off-by: Li Peng <peng.li@intel.com>
* check intel subgroup support for lrn and pooling layer
Signed-off-by: Li Peng <peng.li@intel.com>
* Fix convolution bug if group is greater than 1
Signed-off-by: Li Peng <peng.li@intel.com>
* Set default layer preferableTarget to be DNN_TARGET_CPU
Signed-off-by: Li Peng <peng.li@intel.com>
* Add ocl perf test for convolution
Signed-off-by: Li Peng <peng.li@intel.com>
* Add more ocl accuracy test
Signed-off-by: Li Peng <peng.li@intel.com>
* replace cl_image with ocl::Image2D
Signed-off-by: Li Peng <peng.li@intel.com>
* Fix build failure in elementwise layer
Signed-off-by: Li Peng <peng.li@intel.com>
* use getUMat() to get blob data
Signed-off-by: Li Peng <peng.li@intel.com>
* replace cl_mem handle with ocl::KernelArg
Signed-off-by: Li Peng <peng.li@intel.com>
* dnn(build): don't use C++11, OPENCL_LIBRARIES fix
* dnn(ocl4dnn): remove unused OpenCL kernels
* dnn(ocl4dnn): extract OpenCL code into .cl files
* dnn(ocl4dnn): refine auto-tuning
Defaultly disable auto-tuning, set OPENCV_OCL4DNN_ENABLE_AUTO_TUNING
environment variable to enable it.
Use a set of pre-tuned configs as default config if auto-tuning is disabled.
These configs are tuned for Intel GPU with 48/72 EUs, and for googlenet,
AlexNet, ResNet-50
If default config is not suitable, use the first available kernel config
from the candidates. Candidate priority from high to low is gemm like kernel,
IDLF kernel, basick kernel.
* dnn(ocl4dnn): pooling doesn't use OpenCL subgroups
* dnn(ocl4dnn): fix perf test
OpenCV has default 3sec time limit for each performance test.
Warmup OpenCL backend outside of perf measurement loop.
* use ocl::KernelArg as much as possible
Signed-off-by: Li Peng <peng.li@intel.com>
* dnn(ocl4dnn): fix bias bug for gemm like kernel
* dnn(ocl4dnn): wrap cl_mem into UMat
Signed-off-by: Li Peng <peng.li@intel.com>
* dnn(ocl4dnn): Refine signature of kernel config
- Use more readable string as signture of kernel config
- Don't count device name and vendor in signature string
- Default kernel configurations are tuned for Intel GPU with
24/48/72 EUs, and for googlenet, AlexNet, ResNet-50 net model.
* dnn(ocl4dnn): swap width/height in configuration
* dnn(ocl4dnn): enable configs for Intel OpenCL runtime only
* core: make configuration helper functions accessible from non-core modules
* dnn(ocl4dnn): update kernel auto-tuning behavior
Avoid unwanted creation of directories
* dnn(ocl4dnn): simplify kernel to workaround OpenCL compiler crash
* dnn(ocl4dnn): remove redundant code
* dnn(ocl4dnn): Add more clear message for simd size dismatch.
* dnn(ocl4dnn): add const to const argument
Signed-off-by: Li Peng <peng.li@intel.com>
* dnn(ocl4dnn): force compiler use a specific SIMD size for IDLF kernel
* dnn(ocl4dnn): drop unused tuneLocalSize()
* dnn(ocl4dnn): specify OpenCL queue for Timer and convolve() method
* dnn(ocl4dnn): sanitize file names used for cache
* dnn(perf): enable Network tests with OpenCL
* dnn(ocl4dnn/conv): drop computeGlobalSize()
* dnn(ocl4dnn/conv): drop unused fields
* dnn(ocl4dnn/conv): simplify ctor
* dnn(ocl4dnn/conv): refactor kernelConfig localSize=NULL
* dnn(ocl4dnn/conv): drop unsupported double / untested half types
* dnn(ocl4dnn/conv): drop unused variable
* dnn(ocl4dnn/conv): alignSize/divUp
* dnn(ocl4dnn/conv): use enum values
* dnn(ocl4dnn): drop unused innerproduct variable
Signed-off-by: Li Peng <peng.li@intel.com>
* dnn(ocl4dnn): add an generic function to check cl option support
* dnn(ocl4dnn): run softmax subgroup version kernel first
Signed-off-by: Li Peng <peng.li@intel.com>
In OpenCL code in activations.cl, make the type of floating point
literals to be float. Otherwise the values will be interpreted as
doubles, causing Beignet to have type conversion issues.
fixed problem in concat layer by disabling memory re-use in layers with multiple inputs
trying to fix the tests when Halide is used to run deep nets
another attempt to fix Halide tests
see if the Halide tests will pass with concat layer fusion turned off
trying to fix failures in halide tests; another try
one more experiment to make halide_concat & halide_enet tests pass
continue attempts to fix halide tests
moving on
uncomment parallel concat layer
seemingly fixed failures in Halide tests and re-enabled concat layer fusion; thanks to dkurt for the patch
Main purpose of this namespace is to avoid using of incompatible
binaries that will cause applications crashes.
This additional namespace will not impact "Source code API".
This change allows to maintain ABI checks (with easy filtering out).
* another round of dnn optimization:
* increased malloc alignment across OpenCV from 16 to 64 bytes to make it AVX2 and even AVX-512 friendly
* improved SIMD optimization of pooling layer, optimized average pooling
* cleaned up convolution layer implementation
* made activation layer "attacheable" to all other layers, including fully connected and addition layer.
* fixed bug in the fusion algorithm: "LayerData::consumers" should not be cleared, because it desctibes the topology.
* greatly optimized permutation layer, which improved SSD performance
* parallelized element-wise binary/ternary/... ops (sum, prod, max)
* also, added missing copyrights to many of the layer implementation files
* temporarily disabled (again) the check for intermediate blobs consistency; fixed warnings from various builders
