Ask Your Question

dkurt's profile - activity

2019-01-19 05:07:56 -0500 received badge  Nice Answer (source)
2019-01-18 02:22:24 -0500 answered a question How can I compile OpenCV.js so that it does not require defining the "onRuntimeInitialized" field?

How can I can build an wasm so it does not require me to define "onRuntimeInitialized" WASM is compiled in runtime

2019-01-17 05:10:08 -0500 commented question What are the paid and free packages for unity with module dnn?

Just the first match searching with Google: https://enoxsoftware.com/opencvforunity/

2019-01-17 01:46:42 -0500 commented question Sample tutorial_dnn_android not working without OpenCV Manager

From the tutorial: Install once an appropriate OpenCV manager from unpacked_OpenCV_package/apk to target device.. If you

2019-01-16 02:28:46 -0500 commented question What algorithms are best used for recognizing objects in the street from a video stream

What objects? Maybe you can try to use existing models (cascades or networks) and compare.

2019-01-14 02:47:20 -0500 commented question ArUco detection x-y axis values are extremely different

@yeser, Could you please draw the axes on both frames?

2019-01-14 00:43:30 -0500 commented question readNetFromTensorflow() errors from Mask_RCNN model

@lunasdejavu, According configuration file, you can vary it in feature_extractor field.

2019-01-12 04:42:54 -0500 commented question draw detections when blobFromImages is used

@berak, There is a specification of output for detection_out layers: [batchId, classId, confidence, left, top, right, bo

2019-01-12 03:03:29 -0500 commented question OpenCV (java) : autocapture image in onCameraFrame()

@Akhil Patel, May I ask you clarify it a bit? Do you want to preview camera images without delay? This way you can write

2019-01-11 08:29:44 -0500 commented answer DNN/Tensorflow API works in python but not c++

@richardb, Drawing loop from the code snippet in the question is wrong. Please try to run you model with OpenCV's sample

2019-01-11 08:21:11 -0500 commented answer DNN/Tensorflow API works in python but not c++

Because of wrong postprocessing. Try to use OpenCV's sample for reference: https://github.com/opencv/opencv/blob/master/

2019-01-11 00:29:37 -0500 commented question openCV with openVINO..?

@dilli24, but i cannot use the last demo that using opencv script OpenCV is a part of OpenVINO toolkit so you can

2019-01-11 00:27:48 -0500 commented question openCV with openVINO..?

how can i build opencv with inference engine Do you really need to build OpenCV with IE support? You can use one from Op

2019-01-11 00:05:13 -0500 commented question openCV with openVINO..?

@dilli24, What the output of print(cv.__file__)? It should be something like /path/to/inference_engine_vpu_arm/python/py

2019-01-10 23:14:25 -0500 answered a question DNN/Tensorflow API works in python but not c++

a very similar program in c++ cv.dnn.blobFromImage(img, size=(300, 300), swapRB=True, crop=False) and blobFromIm

2019-01-10 08:33:59 -0500 commented question openCV with openVINO..?

@dilli24, that means you try to use OpenCV which is not from OpenVINO. Check output from print(cv.__file__).

2019-01-10 08:01:43 -0500 commented question openCV with openVINO..?

@supra56, check an installation guide https://software.intel.com/en-us/articles/OpenVINO-Install-RaspberryPI

2019-01-09 02:12:07 -0500 commented question OpenCV crashes after playing a video... help.. please!

@berak, As for me, if a python variable isn't used, it's better to name it correspondingly. In example, _, frame = cap.

2019-01-09 02:11:47 -0500 commented question OpenCV crashes after playing a video... help.. please!

@berak, As for me, if a python variable isn't used, it's better to name it correspondingly. In example, _, frame = cap.r

2019-01-09 00:57:57 -0500 commented question OpenCV crashes after playing a video... help.. please!

@wgb-2019, You need to check that your video is finished. ret is a flag for it.

2019-01-08 04:54:24 -0500 commented question readNetFromTensorflow() errors from Mask_RCNN model

The thing is that currently OpenCV supports only Mask-RCNN network from TensorFlow Object Detection API (see the guide m

2018-12-27 14:05:24 -0500 commented question Dnn.forward crash in Android ndk

@timtsoitt, Never mind. Glad we helped you.

2018-12-27 13:01:34 -0500 commented question Dnn.forward crash in Android ndk

@berak, As for me Error: Assertion failed (inputs[0].size[1] % blobs[0].size[1] == 0) in forward is one of the most regu

2018-12-27 13:00:51 -0500 commented question Dnn.forward crash in Android ndk

@berak, As for me Error: Assertion failed (inputs[0].size[1] % blobs[0].size[1] == 0) in forward is one of the most usua

2018-12-24 05:19:21 -0500 commented question How to debug C++ code in a python program using VS2017?

@LBerger, Python3 is 64-bit?

2018-12-22 05:07:42 -0500 received badge  Good Answer (source)
2018-12-22 04:09:12 -0500 answered a question How to run pretrained model with OpenVINO on RPi

@nikogamulin, using your sample without any changes (except typo in drawing loop at the end) can give me a single licens

2018-12-21 13:48:04 -0500 commented question How to run pretrained model with OpenVINO on RPi

@supra56, https://software.intel.com/articles/OpenVINO-Install-RaspberryPI

2018-12-21 12:57:00 -0500 commented question How to run pretrained model with OpenVINO on RPi

@nikogamulin, please attach the image is used in your sample.

2018-12-21 05:52:15 -0500 commented answer DNN module: switch Torch model between train() and evaluation() modes

@Jaewoo, unfortunately some of our tests are failed with this simple solution. To preserve compatibility we proposed an

2018-12-21 05:51:23 -0500 commented answer DNN module: switch Torch model between train() and evaluation() modes

@Jaewoo, unfortunately some of our tests are failed with this simple solution. To preserve compatibility we proposed an

2018-12-21 05:49:39 -0500 commented answer DNN module: switch Torch model between train() and evaluation() modes

@Jaewoo, unfortunately some of our tests are failed with this simple solution. To preserve compatibility we proposed an

2018-12-21 00:14:16 -0500 commented answer DNN module: switch Torch model between train() and evaluation() modes

@Jaewoo, this PR will fix it: https://github.com/opencv/opencv/pull/13497

2018-12-20 23:46:44 -0500 commented answer DNN module: switch Torch model between train() and evaluation() modes

@Jaewoo, please try the fix mentioned in updated answer. It'd be nice if you can attach some resulting image received fr

2018-12-20 23:44:43 -0500 edited answer DNN module: switch Torch model between train() and evaluation() modes

@Jaewoo, please try the following experiment: 1. Import the model in Torch, generate input, make forward pass and save

2018-12-20 22:28:15 -0500 commented answer DNN module: switch Torch model between train() and evaluation() modes

@Jaewoo, do you really need this? I mean some layers aren't deterministic in training phase (i.e. Dropout). That means f

2018-12-20 06:46:56 -0500 commented question Darknet-Yolo - extract data from Dnn.forward

Take a look at object_detection.py and object_detection.cpp. This Java sample can be also useful: https://github.com/ope

2018-12-20 06:32:22 -0500 commented question Darknet-Yolo - extract data from Dnn.forward

Take a look at object_detection.py and object_detection.cpp

2018-12-20 03:53:53 -0500 edited answer DNN module: switch Torch model between train() and evaluation() modes

@Jaewoo, please try the following experiment: 1. Import the model in Torch, generate input, make forward pass and save

2018-12-20 03:53:03 -0500 answered a question DNN module: switch Torch model between train() and evaluation() modes

@Jaewoo, please try the following experiment: Import the model in Torch, generate input, make forward pass and save re

2018-12-20 03:34:09 -0500 commented question DNN module: switch Torch model between train() and evaluation() modes

@Jaewoo, if you're able to share the model - it'd be the simplest way to reproduce your problem.

2018-12-20 03:31:51 -0500 commented question DNN module: switch Torch model between train() and evaluation() modes

@Jaewoo, I meant just script to generate test data for OpenCV's Torch importer: https://github.com/opencv/opencv_extra/b

2018-12-20 03:24:00 -0500 commented question DNN module: switch Torch model between train() and evaluation() modes

@Jaewoo, At first check that both inputs for train and test phase are similar. Please also check that you use correct me

2018-12-20 03:04:36 -0500 commented question DNN module: switch Torch model between train() and evaluation() modes

@Jaewoo, what do you mean by OpenCV DNN is still using the train mode? How do you test it?

2018-12-20 02:54:56 -0500 commented question DNN module: switch Torch model between train() and evaluation() modes

You need to call evaluation() before Torch's net serialization.

2018-12-18 02:13:40 -0500 commented question OpenCV dnn import quantized model

I guess even TensorFlow will convert the weights in FP32 because it' s just a way to compress model size (read the docs:

2018-12-18 01:24:09 -0500 commented question OpenCV dnn import quantized model

If you mean https://github.com/tensorflow/tensorflow/blob/master/tensorflow/tools/graph_transforms/README.md#quantize_we

2018-12-18 01:15:09 -0500 commented question OpenCV dnn import quantized model

If you mean https://github.com/tensorflow/tensorflow/blob/master/tensorflow/tools/graph_transforms/README.md#quantize_we

2018-12-14 03:09:10 -0500 commented question Image segmentation on Real time

@vineetjai, Could you tell us at least which type of network is used and what problem is solved? Segmentation of such ki

2018-12-12 00:44:13 -0500 commented question How can i load a onnx with opencv?

@velocia, Is it possible to share model.onnx?