Add DNN-based face detection and face recognition into modules/objdetect
* Add DNN-based face detector impl and interface
* Add a sample for DNN-based face detector
* add recog
* add notes
* move samples from samples/cpp to samples/dnn
* add documentation for dnn_face
* add set/get methods for input size, nms & score threshold and topk
* remove the DNN prefix from the face detector and face recognizer
* remove default values in the constructor of impl
* regenerate priors after setting input size
* two filenames for readnet
* Update face.hpp
* Update face_recognize.cpp
* Update face_match.cpp
* Update face.hpp
* Update face_recognize.cpp
* Update face_match.cpp
* Update face_recognize.cpp
* Update dnn_face.markdown
* Update dnn_face.markdown
* Update face.hpp
* Update dnn_face.markdown
* add regression test for face detection
* remove underscore prefix; fix warnings
* add reference & acknowledgement for face detection
* Update dnn_face.markdown
* Update dnn_face.markdown
* Update ts.hpp
* Update test_face.cpp
* Update face_match.cpp
* fix a compile error for python interface; add python examples for face detection and recognition
* Major changes for Vadim's comments:
* Replace class name FaceDetector with FaceDetectorYN in related failes
* Declare local mat before loop in modules/objdetect/src/face_detect.cpp
* Make input image and save flag optional in samples/dnn/face_detect(.cpp, .py)
* Add camera support in samples/dnn/face_detect(.cpp, .py)
* correct file paths for regression test
* fix convertion warnings; remove extra spaces
* update face_recog
* Update dnn_face.markdown
* Fix warnings and errors for the default CI reports:
* Remove trailing white spaces and extra new lines.
* Fix convertion warnings for windows and iOS.
* Add braces around initialization of subobjects.
* Fix warnings and errors for the default CI systems:
* Add prefix 'FR_' for each value name in enum DisType to solve the
redefinition error for iOS compilation; Modify other code accordingly
* Add bookmark '#tutorial_dnn_face' to solve warnings from doxygen
* Correct documentations to solve warnings from doxygen
* update FaceRecognizerSF
* Fix the error for CI to find ONNX models correctly
* add suffix f to float assignments
* add backend & target options for initializing face recognizer
* add checkeq for checking input size and preset size
* update test and threshold
* changes in response to alalek's comments:
* fix typos in samples/dnn/face_match.py
* import numpy before importing cv2
* add documentation to .setInputSize()
* remove extra include in face_recognize.cpp
* fix some bugs
* Update dnn_face.markdown
* update thresholds; remove useless code
* add time suffix to YuNet filename in test
* objdetect: update test code
* Fix gst error handling
* Use the return value instead of the error, which gives no guarantee of being NULL in case of error
* Test err pointer before accessing it
* Remove unreachable code
* videoio(gstreamer): restore check in writer code
Co-authored-by: Alexander Alekhin <alexander.a.alekhin@gmail.com>
Fix ORB integer overflow
* set size_t step to fix integer overflow in ptr0 offset
* added issue_537 test
* minor fix tags, points
* added size_t_step and offset to remove mixed unsigned and signed operations
* features2d: update ORB checks
Co-authored-by: Alexander Alekhin <alexander.a.alekhin@gmail.com>
* dnn: fix unaligned memory access crash on armv7
The getTensorContent function would return a Mat pointing to some
member of a Protobuf-encoded message. Protobuf does not make any
alignment guarantees, which results in a crash on armv7 when loading
models while bit 2 is set in /proc/cpu/alignment (or the relevant
kernel feature for alignment compatibility is disabled). Any read
attempt from the previously unaligned data member would send SIGBUS.
As workaround, this commit makes an aligned copy via existing clone
functionality in getTensorContent. The unsafe copy=false option is
removed. Unfortunately, a rather crude hack in PReLUSubgraph in fact
writes(!) to the Protobuf message. We limit ourselves to fixing the
alignment issues in this commit, and add getTensorContentRefUnaligned
to cover the write case with a safe memcpy. A FIXME marks the issue.
* dnn: reduce amount of .clone() calls
* dnn: update FIXME comment
Co-authored-by: Alexander Alekhin <alexander.a.alekhin@gmail.com>
Make the implementation of optimization in DNN adjustable to different vector sizes with RVV intrinsics.
* Update fastGEMM for multi VLEN.
* Update fastGEMM1T for multi VLEN.
* Update fastDepthwiseConv for multi VLEN.
* Update fastConv for multi VLEN.
* Replace malloc with cv::AutoBuffer.
dnn : int8 quantized layers support in onnx importer
* added quantized layers support in onnx importer
* added more cases in eltwise node, some more checks
* added tests for quantized nodes
* relax thresholds for failed tests, address review comments
* refactoring based on review comments
* added support for unsupported cases and pre-quantized resnet50 test
* relax thresholds due to int8 resize layer
* Prefix global javascript functions with sub-namespaces
* js: handle 'namespace_prefix_override', update filtering
- avoid functions override with same name but different namespace
Co-authored-by: Alexander Alekhin <alexander.a.alekhin@gmail.com>
Add ExpandDims layer of tf_importer.cpp
* Add ExpandDims to tf_importer.
* add -1 expand test case.
* Support different dimensions of input.
* Compatible with 5-dimensional NDHWC data
* Code align
* support 3-dim input.
* 3-dim bug fixed.
* fixing error of code format.
* Add RowVec_8u32f
* Fix build errors in Linux x64 Debug and armeabi-v7a
* Reformat code to make it more clean and conventional
* Optimise with vx_load_expand_q()
Recover pose from different cameras (version 2)
* add recoverPose for two different cameras
* Address review comments from original PR
* Address new review comments
* Rename private api
Co-authored-by: tompollok <tom.pollok@gmail.com>
Co-authored-by: Zane <zane.huang@mail.utoronto.ca>
This submission is used to improve the performance of the inpaint algorithm for 3 channels images(RGB or BGR).
Reason:
The original algorithm implementation did not consider the cache hits.
The loop of channels is outside the core loop, so the perfmance is not very good.
Moving the channel loop inside the core loop can significantly improve cache hits, thereby improving performance.
Performance:
360P, about >= 30% improvement
iphone8P: 5.52ms -> 3.75ms
iphone6s: 14.04ms -> 9.15ms
G-API: Handle reshape for generic case in GExecutor
* Handle reshape for generic case for GExecutor
* Add initResources
* Add tests
* Refactor reshape method
different paddings in cvtColorTwoPlane() for biplane YUV420
* Different paddings support in cvtColorTwoPlane() for biplane YUV420
* Build fix for dispatch case.
* Resoted old behaviour for y.step==uv.step to exclude perf regressions.
Co-authored-by: amir.tulegenov <amir.tulegenov@xperience.ai>
Co-authored-by: Alexander Smorkalov <alexander.smorkalov@xperience.ai>
* feat: OpenCV extension with pure Python modules
* feat: cv2 is now a Python package instead of extension module
Python package cv2 now can handle both Python and C extension modules
properly without additional "subfolders" like "_extra_py_code".
* feat: can call native function from its reimplementation in Python
Add support for YOLOv4x-mish
* backport to 3.4 for supporting yolov4x-mish
* add YOLOv4x-mish test
* address review comments
Co-authored-by: Guo Xu <guoxu@1school.com.cn>
Add CAP_PROP_STREAM_OPEN_TIME
* Added CAP_PROP_STREAM_OPEN_TIME to videoio module - can be used to query the time at which the stream was opened, in seconds since Jan 1 1970 (midnight, UTC). Useful for RTSP and other live video where absolute timestamps are needed. Only applicable to ffmpeg backends
* use nanoseconds instead of seconds to mark the stream open time, and change the cap prop name to CAP_PROP_STREAM_OPEN_TIME_NSEC
* use microseconds for CAP_PROP_STREAM_OPEN_TIME (nanoseconds rolls over too soon, and milliseconds/seconds requires a division)
* fix whitespace issue
Add Normalize subgraph, fix Slice, Mul and Expand
* Add Normalize subgraph, support for starts<0 and axis<0 in Slice, Mul broadcasting in the middle and fix Expand's unsqueeze
* remove todos
* remove range-based for loop
* address review comments
* change >> to > > in template
* fix indexation
* fix expand that does nothing
* support PPSeg model for dnn module
* fixed README for CI
* add test case
* fixed bug
* deal with comments
* rm dnn_model_runner
* update test case
* fixed bug for testcase
* update testcase
`PyObject*` to `std::vector<T>` conversion logic:
- If user passed Numpy Array
- If array is planar and T is a primitive type (doesn't require
constructor call) that matches with the element type of array, then
copy element one by one with the respect of the step between array
elements. If compiler is lucky (or brave enough) copy loop can be
vectorized.
For classes that require constructor calls this path is not
possible, because we can't begin an object lifetime without hacks.
- Otherwise fall-back to general case
- Otherwise - execute the general case:
If PyObject* corresponds to Sequence protocol - iterate over the
sequence elements and invoke the appropriate `pyopencv_to` function.
`std::vector<T>` to `PyObject*` conversion logic:
- If `std::vector<T>` is empty - return empty tuple.
- If `T` has a corresponding `Mat` `DataType` than return
Numpy array instance of the matching `dtype` e.g.
`std::vector<cv::Rect>` is returned as `np.ndarray` of shape `Nx4` and
`dtype=int`.
This branch helps to optimize further evaluations in user code.
- Otherwise - execute the general case:
Construct a tuple of length N = `std::vector::size` and insert
elements one by one.
Unnecessary functions were removed and code was rearranged to allow
compiler select the appropriate conversion function specialization.
[G-API] Extend compileStreaming to support different overloads
* Make different overloads
* Order python compileStreaming overloads
* Fix compileStreaming bug
* Replace
gin -> descr_of
* Set error message
* Fix review comments
* Use macros for pyopencv_to GMetaArgs
* Use GAPI_PROP_RW
* Not split Prims python stuff
* Added exposure and gain props, maximized pixel clk
* removed pixel clock maximization
pixel clock maximization is not suitable for all use cases, so I removed it from PR.
* videoio/gstreamer: Add support for GRAY16_LE.
* videoio/gstreamer: added BGRA/BGRx support
Co-authored-by: Maksim Shabunin <maksim.shabunin@gmail.com>
* VideoCapture timeout set/get
* Common formatting for enum values
* Fix enum values wrongly in videoio.hpp
* Define timeout enum values in public api and align with master
docs(core/ocl): clarify ownership of arguments passed into OpenCL related functions
* docs(core/ocl): clarify ownership in OpenCLExecutionContext::create
Although it is technically true that OpenCLExecutionContext::create
calls `clRetainContext` on its context argument, it is misleading
because it does not increase the reference count overall. Clarify that
the ownership of one reference of the passed context and device is
taken.
* docs(core/ocl): document ownership transfer in ocl::Device::fromHandle
Optimization of DNN using native RISC-V vector intrinsics.
* Use RVV to optimize fastGEMM (FP32) in DNN.
* Use RVV to optimize fastGEMM1T in DNN.
* Use RVV to optimize fastConv in DNN.
* Use RVV to optimize fastDepthwiseConv in DNN.
* Vectorize tails using vl.
* Use "vl" instead of scalar to handle small block in fastConv.
* Fix memory access out of bound in "fastGEMM1T".
* Remove setvl.
* Remove useless initialization.
* Use loop unrolling to handle tail part instead of switch.
[G-API] Support postprocessing for not argmaxed outputs
* Support postprocessing for not argmaxed outputs
* Fix typo
* Add assert
* Remove static cast
* CamelCast to snake_case
* Fix windows warning
* Add static_cast to uint8_t
* Add const to variables
Add Python's test for LSTM layer
* Add Python's test for LSTM layer
* Set different test threshold for FP16 target
* rename test to test_input_3d
Co-authored-by: Julie Bareeva <julia.bareeva@xperience.ai>
