2020-05-23 03:19:51 -0600 | commented question | OpenCV CUDA backend Please use OpenCV 4.3. All devices are supported now. |
2020-05-23 03:19:23 -0600 | commented answer | OpenCV CUDA backend All devices are supported since OpenCV 4.3. |
2020-05-23 03:18:09 -0600 | commented question | CUDA DNN initialization when changing in batch size You haven't done anything wrong. That's just the way it is. Changing the input shape causes reinitialization. The only w |
2020-05-23 03:14:45 -0600 | commented question | Building DNN module with cuDNN backend @rgov The issue linked in the post is outdated. All compute capabilities are now supported by the CUDA backend. The mess |
2020-04-26 02:44:52 -0600 | commented question | How to best free up DNN and GPU resources when finished? net = cv::dnn::Net() should do the trick. |
2020-04-11 01:22:06 -0600 | commented question | OpenCV::DNN doesn't support conv1d You should open a feature request on GitHub. |
2020-03-10 05:20:27 -0600 | commented question | switch graphic card in Cuda https://github.com/opencv/opencv/issues/16725 |
2020-03-06 00:51:32 -0600 | commented question | Can opencv and Tensorflow use GPU simultaneously? @Sshwan what GPU you have? You can reduce memory consumption by using the DNN_TARGET_CUDA_FP16 target. You will reduce t |
2020-02-29 11:41:52 -0600 | commented question | how could I change memory layout from hwc to chw ? You can reshape (H, W, C) to (HW, C) and then perform a transpose. Transpose would give (C, HW) which can then be reshap |
2020-01-03 23:42:14 -0600 | commented answer | How to use cudacodec + blobFromImage @cudawarped It's there in my list. |
2020-01-03 22:43:45 -0600 | commented answer | How to use cudacodec + blobFromImage @cudawarped I have added it to my list. EDIT: It was already there :) |
2020-01-03 22:42:37 -0600 | commented answer | How to use cudacodec + blobFromImage @cudawarped I have added it to my list. |
2020-01-02 11:36:57 -0600 | commented question | opencv dnn shows error when use resnet50.cfg from darknet This configuration is broken in OpenCV DNN. You can track the bug here (GitHub Issue 16259). |
2019-12-30 22:42:56 -0600 | commented answer | How to use cudacodec + blobFromImage @VectorVP Yes, it is planned. It's just 1 hour work once GpuMat gets support for arbitrary dimensions. |
2019-12-30 22:24:18 -0600 | commented answer | How to use cudacodec + blobFromImage @VectorVP Yes, it is planned. |
2019-12-30 11:45:39 -0600 | commented answer | How to use cudacodec + blobFromImage There is no GpuMat support yet because it does not support arbitrary dimensions. It's possible to allow single image inf |
2019-12-30 10:49:09 -0600 | commented answer | How to use cudacodec + blobFromImage There is no GpuMat support yet because it does not support arbitrary dimensions. It's possible to allow single image inf |
2019-12-29 04:20:42 -0600 | commented question | dnn with cuda on 4.20 Build and usage instructions can be found here. Example code to use OpenCV DNN CUDA for YOLOv3. Note that you have to se |
2019-12-29 04:18:52 -0600 | commented question | dnn with cuda on 4.20 Build and usage instructions can be found here. Example code to use OpenCV DNN CUDA for YOLOv3. Note that you have to se |
2019-12-29 04:17:58 -0600 | commented question | dnn with cuda on 4.20 Build and usage instructions can be found here. |
2019-12-29 04:17:04 -0600 | commented answer | dnn with cuda on 4.20 1080 Ti doesn't have good half-precision throughput. DNN_TARGET_CUDA (which is the FP32 target) will be faster than DNN_ |
2019-12-26 11:31:52 -0600 | commented question | openCV 4.2 with CUDA 10.2 OpenCV cannot train models but it can perform inference slightly faster than darknet. The heavier the model, the larger |
2019-12-26 11:31:41 -0600 | commented question | openCV 4.2 with CUDA 10.2 OpenCV cannot train models but it can perform inference slightly faster than darknet. The heavier the model, the larger |
2019-12-26 11:30:51 -0600 | commented question | openCV 4.2 with CUDA 10.2 OpenCV cannot train models but it can perform inference slightly faster than darknet. The heavier the model, the larger |
2019-12-26 03:24:15 -0600 | commented question | openCV 4.2 with CUDA 10.2 OpenCV has got a CUDA backend for the DNN module. Please have a look at this benchmark. If you tell me the model you pla |
2019-12-26 03:23:47 -0600 | commented question | openCV 4.2 with CUDA 10.2 OpenCV has got a CUDA backend for the DNN module. Please have a look at this benchmark. If you tell me the model you pla |
2019-12-26 03:18:31 -0600 | commented question | openCV 4.2 with CUDA 10.2 OpenCV has got a CUDA backend for the DNN module. Please have a look at this benchmark. If you tell me the model you pla |
2019-12-26 03:16:29 -0600 | commented answer | openCV 4.2 with CUDA 10.2 I am using 10.2 on Ubuntu 18.04 and it works great. |
2019-12-22 05:51:24 -0600 | commented answer | CUDA backend for DNN module requires CC 5.3 or higher @cudawarped No, it provides both options separately. There is an open PR which makes half-precision optional so that it |
2019-12-21 07:53:07 -0600 | commented answer | CUDA backend for DNN module requires CC 5.3 or higher It's 5.3 according to this table. |
2019-12-07 10:21:57 -0600 | commented question | CUDA backend for DNN module requires CC 5.3 or higher You can compile OpenCV with CUDA support.You will have to disable the CUDA backend in the DNN module by unmarking OPENCV |
2019-12-07 09:10:20 -0600 | commented answer | How to run OpenCV DNN on NVidia GPU @muz Assuming that you have built the master (because the CUDA backend is not yet in a release), you have to set backend |
2019-12-02 05:58:59 -0600 | commented answer | How to run OpenCV DNN on NVidia GPU There is a CUDA backend in OpenCV DNN module now which is much faster than the OpenCL backend. |
2019-12-02 05:56:44 -0600 | commented question | CUDA backend for DNN module requires CC 5.3 or higher You can compile OpenCV with CUDA support.You will have to disable the CUDA backend in the DNN module by unmarking OPENCV |
2019-11-15 01:54:39 -0600 | edited answer | Error compiling openCV 4.1.2 with CUDA 9.0 in Ubuntu 16.04 CUDA 10.0 added native vector arithmetic operators for the __half2 datatype. The CUDA backend for OpenCV DNN makes use o |
2019-11-15 01:52:06 -0600 | answered a question | Error compiling openCV 4.1.2 with CUDA 9.0 in Ubuntu 16.04 CUDA 10.0 added native vector arithmetic operators for the __half2 datatype. The CUDA backend for OpenCV DNN has merged |
2019-10-27 23:39:59 -0600 | commented question | Building OpenCV 4 with CUDA Yes, it adds a new backend to the DNN module which accelerates DNN inference on CUDA capable cards dramatically. |
2019-10-27 23:37:13 -0600 | commented question | Does opencv_dnn use gpu? Download the master branch as a ZIP (you will see a GREEN button saying "clone or download" in the OpenCV repository). E |
2019-10-27 23:34:30 -0600 | commented question | Does opencv_dnn use gpu? @redhwan The CUDA backend hasn't been released yet. It's available on the master branch (where stuff for the next releas |
2019-10-27 21:31:58 -0600 | received badge | ● Nice Answer (source) |
2019-10-27 13:09:33 -0600 | commented question | Building OpenCV 4 with CUDA FYI: OpenCV has had CUDA support for a few years. The linked PR only adds CUDA support to the OpenCV DNN module. |
2019-10-27 13:06:34 -0600 | commented question | Building OpenCV 4 with CUDA FYI: OpenCV has had CUDA support for years. The linked PR only adds CUDA support to the OpenCV DNN module. |
2019-10-27 12:55:11 -0600 | commented question | Does opencv_dnn use gpu? @Erik Langskjegg OPENCV_DNN_CUDA is also a CMake option just like the other two. |
2019-10-27 12:53:27 -0600 | edited answer | Building OpenCV 4 with CUDA OPENCV_DNN_CUDA is also a CMake option. You can set it using -DOPENCV_DNN_CUDA=ON. |
2019-10-27 12:52:07 -0600 | answered a question | Building OpenCV 4 with CUDA `OPENCV_DNN_CUDA |
2019-10-27 12:50:47 -0600 | commented question | Does opencv_dnn use gpu? @Erik Langskjegg They are all CMake options. The OPENCV_DNN_CUDA option be present amongst others by default. It's not m |
2019-10-27 12:49:59 -0600 | commented question | Does opencv_dnn use gpu? @Erik Langskjegg The OPENCV_DNN_CUDA option be present amongst others by default. It's not marked by default but it shou |
2019-10-27 07:01:32 -0600 | commented question | Does opencv_dnn use gpu? @redhwan your question was posted before the CUDA support was merged into master. I preassume that your OpenCV build doe |
2019-10-27 00:49:34 -0600 | commented answer | Does opencv_dnn use gpu? The CUDA support is currently available in the master branch only. Assuming that you have built the master branch, you h |
2019-10-26 22:59:10 -0600 | commented question | OpenCV dnn, OpenCL and Nvidia graphic cards @holger It's in master now. |