'solved - OpenCV DNN YOLO V4 with CUDA in Python is 5x Faster than the same code in C++

TL;DR; it was a stupid bug in my code. Not a cuda problem at all.

I'm trying to compare the performance of the following code:

frames = ...
for i in range(2000):
    frame = frames[i % 4]
    model.detect(frame, .2, .4)

and its counterpart in C++:

frames = ...
for(int i = 0; i < 2000; ++i) {
    const cv::Mat & frame = frames[i % 4];
    model.detect(frame, classIds, confidences, boxes, .2, .4);
}

In C++, the backend for CUDA is set by:

net.setPreferableBackend(cv::dnn::DNN_BACKEND_CUDA);
net.setPreferableTarget(cv::dnn::DNN_TARGET_CUDA_FP16);

In Python, the backend for CUDA is set by:

net.setPreferableBackend(cv2.dnn.DNN_BACKEND_CUDA)
net.setPreferableTarget(cv2.dnn.DNN_TARGET_CUDA_FP16)

Surprisingly, I've found that the Python program runs very much faster, performing at 300 FPS meanwhile the equivalent C++ program runs at 60 FPS. In both executions, CUDA is enabled.

The code and details about my environment can be checked in a repository I created for this analysis.

It is noteworthy that running the programs again using only the CPU (CUDA disabled) results in 60 FPS for either C++ and Python version. This let me think that OpenCV is not actually using CUDA for the cv::dnn API. Indeed, the C++ program consumes 6% of the GPU meanwhile the Python program uses 67% of the GPU.

Do someone have already find a similar scenario like that? Or at other hand, what I'm doing wrong here in the C++ side when running the code with CUDA?

EDIT: The output of cv::getBuildInformation() is:

General configuration for OpenCV 4.5.3 =====================================
  Version control:               4.5.3

  Extra modules:
    Location (extra):            /home/doleron/opencv_build/opencv_contrib/modules
    Version control (extra):     4.5.3

  Platform:
    Timestamp:                   2022-01-16T16:15:55Z
    Host:                        Linux 5.11.0-46-generic x86_64
    CMake:                       3.16.3
    CMake generator:             Unix Makefiles
    CMake build tool:            /usr/bin/make
    Configuration:               RELEASE

  CPU/HW features:
    Baseline:                    SSE SSE2 SSE3
      requested:                 SSE3
    Dispatched code generation:  SSE4_1 SSE4_2 FP16 AVX AVX2 AVX512_SKX
      requested:                 SSE4_1 SSE4_2 AVX FP16 AVX2 AVX512_SKX
      SSE4_1 (15 files):         + SSSE3 SSE4_1
      SSE4_2 (1 files):          + SSSE3 SSE4_1 POPCNT SSE4_2
      FP16 (0 files):            + SSSE3 SSE4_1 POPCNT SSE4_2 FP16 AVX
      AVX (4 files):             + SSSE3 SSE4_1 POPCNT SSE4_2 AVX
      AVX2 (29 files):           + SSSE3 SSE4_1 POPCNT SSE4_2 FP16 FMA3 AVX AVX2
      AVX512_SKX (4 files):      + SSSE3 SSE4_1 POPCNT SSE4_2 FP16 FMA3 AVX AVX2 AVX_512F AVX512_COMMON AVX512_SKX

  C/C++:
    Built as dynamic libs?:      YES
    C++ standard:                11
    C++ Compiler:                /usr/bin/c++  (ver 9.3.0)
    C++ flags (Release):         -fsigned-char -ffast-math -W -Wall -Werror=return-type -Werror=non-virtual-dtor -Werror=address -Werror=sequence-point -Wformat -Werror=format-security -Wmissing-declarations -Wundef -Winit-self -Wpointer-arith -Wshadow -Wsign-promo -Wuninitialized -Wsuggest-override -Wno-delete-non-virtual-dtor -Wno-comment -Wimplicit-fallthrough=3 -Wno-strict-overflow -fdiagnostics-show-option -Wno-long-long -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections  -msse -msse2 -msse3 -fvisibility=hidden -fvisibility-inlines-hidden -O3 -DNDEBUG  -DNDEBUG
    C++ flags (Debug):           -fsigned-char -ffast-math -W -Wall -Werror=return-type -Werror=non-virtual-dtor -Werror=address -Werror=sequence-point -Wformat -Werror=format-security -Wmissing-declarations -Wundef -Winit-self -Wpointer-arith -Wshadow -Wsign-promo -Wuninitialized -Wsuggest-override -Wno-delete-non-virtual-dtor -Wno-comment -Wimplicit-fallthrough=3 -Wno-strict-overflow -fdiagnostics-show-option -Wno-long-long -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections  -msse -msse2 -msse3 -fvisibility=hidden -fvisibility-inlines-hidden -g  -O0 -DDEBUG -D_DEBUG
    C Compiler:                  /usr/bin/cc
    C flags (Release):           -fsigned-char -ffast-math -W -Wall -Werror=return-type -Werror=address -Werror=sequence-point -Wformat -Werror=format-security -Wmissing-declarations -Wmissing-prototypes -Wstrict-prototypes -Wundef -Winit-self -Wpointer-arith -Wshadow -Wuninitialized -Wno-comment -Wimplicit-fallthrough=3 -Wno-strict-overflow -fdiagnostics-show-option -Wno-long-long -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections  -msse -msse2 -msse3 -fvisibility=hidden -O3 -DNDEBUG  -DNDEBUG
    C flags (Debug):             -fsigned-char -ffast-math -W -Wall -Werror=return-type -Werror=address -Werror=sequence-point -Wformat -Werror=format-security -Wmissing-declarations -Wmissing-prototypes -Wstrict-prototypes -Wundef -Winit-self -Wpointer-arith -Wshadow -Wuninitialized -Wno-comment -Wimplicit-fallthrough=3 -Wno-strict-overflow -fdiagnostics-show-option -Wno-long-long -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections  -msse -msse2 -msse3 -fvisibility=hidden -g  -O0 -DDEBUG -D_DEBUG
    Linker flags (Release):      -Wl,--exclude-libs,libippicv.a -Wl,--exclude-libs,libippiw.a   -Wl,--gc-sections -Wl,--as-needed  
    Linker flags (Debug):        -Wl,--exclude-libs,libippicv.a -Wl,--exclude-libs,libippiw.a   -Wl,--gc-sections -Wl,--as-needed  
    ccache:                      NO
    Precompiled headers:         NO
    Extra dependencies:          m pthread cudart_static dl rt nppc nppial nppicc nppidei nppif nppig nppim nppist nppisu nppitc npps cublas cudnn cufft -L/usr/local/cuda/lib64 -L/usr/lib/x86_64-linux-gnu
    3rdparty dependencies:

  OpenCV modules:
    To be built:                 aruco barcode bgsegm bioinspired calib3d ccalib core cudaarithm cudabgsegm cudacodec cudafeatures2d cudafilters cudaimgproc cudalegacy cudaobjdetect cudaoptflow cudastereo cudawarping cudev datasets dnn dnn_objdetect dnn_superres dpm face features2d flann freetype fuzzy gapi hfs highgui img_hash imgcodecs imgproc intensity_transform line_descriptor mcc ml objdetect optflow phase_unwrapping photo plot python3 quality rapid reg rgbd saliency shape stereo stitching structured_light superres surface_matching text tracking video videoio videostab wechat_qrcode xfeatures2d ximgproc xobjdetect xphoto
    Disabled:                    world
    Disabled by dependency:      -
    Unavailable:                 alphamat cvv hdf java julia matlab ovis python2 sfm ts viz
    Applications:                apps
    Documentation:               NO
    Non-free algorithms:         YES

  GUI: 
    GTK+:                        YES (ver 3.24.20)
      GThread :                  YES (ver 2.64.6)
      GtkGlExt:                  NO
    VTK support:                 NO

  Media I/O: 
    ZLib:                        /usr/lib/x86_64-linux-gnu/libz.so (ver 1.2.11)
    JPEG:                        /usr/lib/x86_64-linux-gnu/libjpeg.so (ver 80)
    WEBP:                        build (ver encoder: 0x020f)
    PNG:                         /usr/lib/x86_64-linux-gnu/libpng.so (ver 1.6.37)
    TIFF:                        /usr/lib/x86_64-linux-gnu/libtiff.so (ver 42 / 4.1.0)
    JPEG 2000:                   build (ver 2.4.0)
    OpenEXR:                     build (ver 2.3.0)
    HDR:                         YES
    SUNRASTER:                   YES
    PXM:                         YES
    PFM:                         YES

  Video I/O:
    DC1394:                      YES (2.2.5)
    FFMPEG:                      YES
      avcodec:                   YES (58.54.100)
      avformat:                  YES (58.29.100)
      avutil:                    YES (56.31.100)
      swscale:                   YES (5.5.100)
      avresample:                NO
    v4l/v4l2:                    YES (linux/videodev2.h)

  Parallel framework:            TBB (ver 2020.2 interface 11102)

  Trace:                         YES (with Intel ITT)

  Other third-party libraries:
    Intel IPP:                   2020.0.0 Gold [2020.0.0]
           at:                   /home/doleron/opencv_build/opencv/build/3rdparty/ippicv/ippicv_lnx/icv
    Intel IPP IW:                sources (2020.0.0)
              at:                /home/doleron/opencv_build/opencv/build/3rdparty/ippicv/ippicv_lnx/iw
    VA:                          NO
    Lapack:                      NO
    Eigen:                       NO
    Custom HAL:                  NO
    Protobuf:                    build (3.5.1)

  NVIDIA CUDA:                   YES (ver 11.6, CUFFT CUBLAS FAST_MATH)
    NVIDIA GPU arch:             75
    NVIDIA PTX archs:

  cuDNN:                         YES (ver 8.3.2)

  OpenCL:                        YES (no extra features)
    Include path:                /home/doleron/opencv_build/opencv/3rdparty/include/opencl/1.2
    Link libraries:              Dynamic load

  Python 3:
    Interpreter:                 /usr/bin/python3 (ver 3.8.10)
    Libraries:                   /usr/lib/x86_64-linux-gnu/libpython3.8.so (ver 3.8.10)
    numpy:                       /usr/lib/python3/dist-packages/numpy/core/include (ver 1.17.4)
    install path:                lib/python3.8/dist-packages/cv2/python-3.8

  Python (for build):            /usr/bin/python3

  Java:                          
    ant:                         NO
    JNI:                         NO
    Java wrappers:               NO
    Java tests:                  NO

  Install to:                    /usr/local
-----------------------------------------------------------------

python c++opencv yolov4

Solution 1:^[1]

hmm..do a comparison based on data in your repo (C++ version), (1) and (2) below are pretty close, somehow it seems that CUDA does not work properly, both might be running on CPU?

(1) C++/CUDA: real = 0m33,179s; user = 6m14,921s; sys = 0m6,942s

(2) C++/CPU: real = 0m34,341s; user = 6m19,379s; sys = 0m7,908s

Sources

This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.

Source: Stack Overflow

Solution	Source
Solution 1	eleven

'solved - OpenCV DNN YOLO V4 with CUDA in Python is 5x Faster than the same code in C++

Solution 1:[1]

Sources

Related Questions

Solution 1:^[1]