[GSoC] High Level API and Samples for Scene Text Detection and Recognition by wenqingzhang-gordon · Pull Request #17570 · opencv/opencv

wenqingzhang-gordon · 2020-06-17T11:11:18Z

Merge with extra: opencv/opencv_extra#773
High-Level API and Samples for Scene Text Detection and Recognition
This is my project in GSoC 2020: OpenCV Text/Digit detection & recognition projects.

Short Video: https://drive.google.com/file/d/1IlGpRRhPCifC9TRzuhq0_G1P6MkP33BJ/view?usp=sharing

For more information:
https://github.com/HannibalAPE/opencv/blob/text_det_recog_demo/doc/tutorials/dnn/dnn_text_spotting/dnn_text_spotting.markdown

TODO LIST:

Scene Text Recognition:
Samples
High-Level API (CRNN)
Scene Text Detection:
Samples
High-Level API (DB & EAST)
Scene Text Spotting:
Samples
Document:
Tutorials

Pull Request Readiness Checklist

I agree to contribute to the project under OpenCV (Apache 2) License.
To the best of my knowledge, the proposed patch is not based on a code under GPL or other license that is incompatible with OpenCV
The PR is proposed to proper branch
There is reference to original bug report and related work
There is accuracy test, performance test and test data in opencv_extra repository, if applicable
But the patch to opencv_extra does not have the same branch name. see as follows.
The feature is well documented and sample code can be built with the project CMake

opencv_extra=text_det_rec

build_image:Custom=centos:7
buildworker:Custom=linux-1

berak · 2020-06-17T12:01:54Z

samples/dnn/scene_text_recognition.cpp

+        }
+        if (maxLoc > 0) {
+            char currentChar = vocabulary[maxLoc - 1];
+            if (currentChar != decodeSeq[-1])


decodeSeq[-1]

-1 is illegal, maybe use decodeSeq.back() instead ?

Fix it.
I got the right result with index -1, so I thought a negative index may be supported now. :(
Thank you.

berak · 2020-06-17T12:13:07Z

samples/dnn/scene_text_detection.cpp

+        "{ help  h                          | | Print help message. }"
+        "{ inputImage i                     | | Path to an input image. Skip this argument to capture frames from a camera. }"
+        "{ device d                         | 0 | camera device number. }"
+        "{ modelPath mp                     | | Path to a binary .onnx file contains trained DB detector model.}"


just curious -- is there a pretrained onnx model (link) ?

Maybe I can upload models to Google Drive.
I am not sure about it, and I will ask my mentor.

bhack · 2020-07-13T14:29:27Z

modules/dnn/test/test_model.cpp


+TEST_P(Test_Model, SceneTextRec)
+{
+    std::string imgPath = _tf("welcome.png");


Where is this file?

I have already put the data into opencv/opencv_extra, but it has not been merged yet.
opencv/opencv_extra#773
The name of image has been changed into "text_rec_test.png", and I will push again when the data is ready.
Thanks for your review.

Please push the new name. You don't need to wait the merge. See this step in the build pipeline.

Thank you.
Sadly, I still get an error in "Linux x64 Debug", but I find that the test_dnn in "Linux x64" passed. I am not sure about the difference between these two tests.

Is there any detailed log information containing which line throws out the error.
I can only get "error: (-215: Assertion failed) dims <= 2 in function 'at' thrown in the test body." in https://pullrequest.opencv.org/buildbot/builders/precommit_linux64_no_opt/builds/24676/steps/test_dnn/logs/stdio

I think I have tested the API successfully, and you can see more information in https://github.com/HannibalAPE/opencv/blob/text_det_recog_demo/doc/tutorials/dnn/dnn_scene_text_det_and_rec/scene_text_recognition.markdown

I still get an error in "Linux x64 Debug", but I find that the test_dnn in "Linux x64" passed. I am not sure about the difference between these two tests.

Obviously, it is "Debug" mode (with extra checks).

@alalek Thanks for your reminder.

bhack · 2020-07-13T15:08:28Z

@vpisarev I suppose that we could not have a single PR looking at the TODO list in the description.

samples/dnn/scene_text_detection.cpp

vpisarev · 2020-07-27T15:36:45Z

@HannibalAPE, I'm now trying to run the code. None of the models you provided can be read with the version of OpenCV that is in text_det_recog_demo. BTW, hash sum for onnx/models/DB_IC15_resnet50.onnx does not match as well when I run download_models.py script. Could you please check that the models are imported correctly. Make sure that in CMake the version of protobuf from OpenCV is used, not the system version.

wenqingzhang-gordon · 2020-07-27T16:38:49Z

@HannibalAPE, I'm now trying to run the code. None of the models you provided can be read with the version of OpenCV that is in text_det_recog_demo. BTW, hash sum for onnx/models/DB_IC15_resnet50.onnx does not match as well when I run download_models.py script. Could you please check that the models are imported correctly. Make sure that in CMake the version of protobuf from OpenCV is used, not the system version.

@vpisarev

I have updated the sha of DB_IC15_resnet50.onnx. I forgot to update it when I updated the model. sry.
I compile it again, and I am sure that my models can be imported.
I checked the protobuf version in the CMAKE.

  Other third-party libraries:
    Lapack:                      YES (/usr/lib/x86_64-linux-gnu/liblapack.so /usr/lib/x86_64-linux-gnu/libcblas.so /usr/lib/x86_64-linux-gnu/libatlas.so)
    Eigen:                       NO
    Custom HAL:                  NO
    Protobuf:                    build (3.5.1)

Is it not right?
When I compile it, only the ippicv_2020_lnx_intel64_20191018_general.tgz downloading is failed. Dose it matter?

wenqingzhang-gordon · 2020-07-28T12:11:11Z

@vpisarev Hi Vadim, I have tested it on other Linux-based systems (e.g. Ubuntu 18.04 and 16.04). On both of them, my models can be imported and output the right results. Can you provide some information about your issue, then I can reproduce it.

vpisarev · 2020-07-29T22:10:42Z

@HannibalAPE, thank you. I've downloaded the latest DB_IC15_resnet50.onnx and it works well. It's noticeably slower than EAST detector, but the results are definitely better!

However, with the model DB_TD500_resnet50.onnx it still crashes with the following message:

[ERROR:0] global /Users/vpisarev/work/opencv/modules/dnn/src/dnn.cpp (3272) getLayerShapesRecursively OPENCV/DNN: [Reshape]:(712): getMemoryShapes() throws exception. inputs=1 outputs=1/1 blobs=0
[ERROR:0] global /Users/vpisarev/work/opencv/modules/dnn/src/dnn.cpp (3275) getLayerShapesRecursively     input[0] = [ 1 1 736 1280 ]
[ERROR:0] global /Users/vpisarev/work/opencv/modules/dnn/src/dnn.cpp (3279) getLayerShapesRecursively     output[0] = [ ]
[ERROR:0] global /Users/vpisarev/work/opencv/modules/dnn/src/dnn.cpp (3285) getLayerShapesRecursively Exception message: OpenCV(4.4.0-pre) /Users/vpisarev/work/opencv/modules/dnn/src/layers/reshape_layer.cpp:113: error: (-215:Assertion failed) total(srcShape, srcRange.start, srcRange.end) == maskTotal in function 'computeShapeByReshapeMask'

libc++abi.dylib: terminating with uncaught exception of type cv::Exception: OpenCV(4.4.0-pre) /Users/vpisarev/work/opencv/modules/dnn/src/layers/reshape_layer.cpp:113: error: (-215:Assertion failed) total(srcShape, srcRange.start, srcRange.end) == maskTotal in function 'computeShapeByReshapeMask'

Abort trap: 6

I'm using macOS 10.15.6, xcode 11.6. Protobuf is 3.5.1. Will try it on Linux tomorrow or on Friday.

Some other complains:

As I said, the speed is not that good. I tried to play with "--inputWidth" and "--inputHeight" parameters.

First of all, it's inconvenient that you need to set both these fields. I would compute "height" out of "width" for example if just "width" is set and vice versa.
Ok, when I specify both parameters, inputWidth, inputHeight, the app crashes again (with DB_IC15_resnet50.onnx model):

V16M:Release vpisarev$ ./example_dnn_scene_text_detection --mp=DB_IC15_resnet50.onnx --inputWidth=640 --inputHeight=368
[ERROR:0] global /Users/vpisarev/work/opencv/modules/dnn/src/dnn.cpp (3272) getLayerShapesRecursively OPENCV/DNN: [Eltwise]:(551): getMemoryShapes() throws exception. inputs=2 outputs=0/1 blobs=0
[ERROR:0] global /Users/vpisarev/work/opencv/modules/dnn/src/dnn.cpp (3275) getLayerShapesRecursively     input[0] = [ 1 256 24 40 ]
[ERROR:0] global /Users/vpisarev/work/opencv/modules/dnn/src/dnn.cpp (3275) getLayerShapesRecursively     input[1] = [ 1 256 23 40 ]
[ERROR:0] global /Users/vpisarev/work/opencv/modules/dnn/src/dnn.cpp (3285) getLayerShapesRecursively Exception message: OpenCV(4.4.0-pre) /Users/vpisarev/work/opencv/modules/dnn/src/layers/eltwise_layer.cpp:216: error: (-215:Assertion failed) inputs[0][j] == inputs[i][j] in function 'getMemoryShapes'

libc++abi.dylib: terminating with uncaught exception of type cv::Exception: OpenCV(4.4.0-pre) /Users/vpisarev/work/opencv/modules/dnn/src/layers/eltwise_layer.cpp:216: error: (-215:Assertion failed) inputs[0][j] == inputs[i][j] in function 'getMemoryShapes'

can you please check if inputWidth/inputHeight work at all?

==

The text recognition sample says it supports live video capture from camera, but if I run it without image, it complains that the image is not set. From the code I can conclude that it does not support live video capture. Can you modify scene_text_detection sample to support recognition as well? It would be very useful demonstration on how to use detection and recognition together.

wenqingzhang-gordon · 2020-07-30T00:55:15Z

However, with the model DB_TD500_resnet50.onnx it still crashes with the following message:

@vpisarev It may be caused by the wrong input size. For DB_TD500_resnet50.onnx, the image shape should be set to 736x736, which is mentioned in both the tutorial and the sample.

"--inputWidth" and "--inputHeight"

These two parameters are actually prepared for different models, because these models are trained on different benchmarks.
The performance is related to the shape of training samples.

when I specify both parameters, inputWidth, inputHeight, the app crashes again

Currently, it only supports a predefined shape. I will try to update it to support dynamic shape. But I think there will be a drop in accuracy, because the inference shape is not consistent with the training shape.

It would be very useful demonstration on how to use detection and recognition together.

This sample and some new tutorials will be pushed in this weak.

alalek

Thank you for contribution!

alalek · 2020-09-03T19:44:29Z

doc/tutorials/dnn/dnn_text_spotting/dnn_text_spotting.markdown

+
+```cpp
+#include <iostream>
+#include <fstream>


Create standalone .cpp file for example code and embed snippets into documentation only.

You can adopt sample file below or create new one.

doc/tutorials/dnn/dnn_text_spotting/dnn_text_spotting.markdown

modules/dnn/include/opencv2/dnn/dnn.hpp

alalek · 2020-09-05T17:33:36Z

doc/tutorials/dnn/dnn_text_spotting/detect_test1.png
doc/tutorials/dnn/dnn_text_spotting/detect_test2.png

PNG lossless format is not necessary here (files are large for real world images). Try to reduce image size using the .jpg format

alalek · 2020-09-05T17:35:08Z

doc/tutorials/dnn/dnn_text_spotting/text_recognition.cpp

Sample code for tutorials should to into samples/cpp/tutorial_code/...

alalek

Looks good to me!

Please take a look on left comments

doc/tutorials/dnn/table_of_content_dnn.markdown

samples/dnn/scene_text_detection.cpp

alalek · 2020-09-05T17:40:42Z

samples/dnn/scene_text_detection.cpp

+    } else {
+        // Open an image file
+        CV_Assert(parser.has("inputImage"));
+        Mat frame = imread(parser.get<String>("inputImage"));


Please use samples::findFile() for better file searching experience.

I have updated all imread with samples::findFile()

alalek · 2020-09-05T17:43:53Z

@dkurt Please take look on public API proposals.

dkurt · 2020-09-07T05:42:16Z

modules/dnn/include/opencv2/dnn/dnn.hpp

+         /**
+          * @brief Given the @p input frame, create input blob, run net and return recognition result.
+          * @param[in] frame: The input image.
+          * @param[in] decodeType: The decoding method of translating the network output into string.


Which are possible values?

It only supports "CTC-greedy" now, and I will add more decoding methods in the future, such as Beam Searching.

User should know the options so it must be documented or we can add enum

dkurt · 2020-09-07T05:42:51Z

modules/dnn/include/opencv2/dnn/dnn.hpp

+          * @param[in] binThresh: The threshold of the binary map.
+          * @param[in] polyThresh: The threshold of text polygons.
+          * @param[in] unclipRatio: The unclip ratio of the detected text region.
+          * @param[in] maxCandidates: The max number of the output polygons.


Why we need to limit this number?

We can get good detection results with these parameters, for more information you can refer to the paper and code.

In some difficult cases, the output map of the network is full of small noise, and maxCandidates can avoid wasting inference time.

wenqingzhang-gordon · 2020-11-19T11:46:47Z

@alalek
I have no idea about how to update the struct Voc without modification to struct Impl.
I tried to define TextRecognition::Voc inherited from Model::Impl, but I do not know how to apply "Bridge" design pattern here.
Do you have any suggestions about it? Is there any example?
Thank you.

alalek

@HannibalAPE Thank you for contribution!
I will update introduced public APIs and push them here till the end of this week.

Update: Done

wenqingzhang-gordon

@alalek Thank you! I learn a lot with your help.

wenqingzhang-gordon · 2020-11-20T02:44:38Z

samples/dnn/scene_text_recognition.cpp

    CV_Assert(!vocPath.empty());
    std::ifstream vocFile;
-    vocFile.open(vocPath);
+    vocFile.open(samples::findFile(vocPath));


I wonder whether I need to add samples::findFile() whenever I open files?
Do I need to change TextRecognitionModel recognizer(modelPath) into TextRecognitionModel recognizer(samples::findFile(modelPath)) ?
Does it slow the speed?

samples::findFile() helps to use file by name (instead of full path) from <opencv>/samples/data location.

Model file is not stored there (and no plans to put it there due its size).

alalek

@HannibalAPE

I pushed updated public API for text detection / recognition tasks. Please take a look and check samples / documentation (probably I miss something).

alalek · 2020-11-21T16:06:39Z

modules/dnn/src/model.cpp

+    {
+        CV_TRACE_FUNCTION();
+        std::vector< std::vector<Point> > contours = detectTextContours(frame);
+        confidences = std::vector<float>(contours.size(), 1.0f);


Is there any confidences scoring in DB detection algo?

https://github.com/HannibalAPE/opencv/blob/f59aa6d4ae76e08f199e56ae94235295e0380076/modules/dnn/src/model.cpp#L972

You can regard the return of contourScore() as a kind of confidence, but it is only used to filter some bad detection results. It is not the same meaning as those in general object detection algorithms.

modules/dnn/include/opencv2/dnn/dnn.hpp

wenqingzhang-gordon · 2020-11-24T15:16:31Z

@alalek Since the onnx importer has supported models with dynamic input shape, I will try to generate new models of DB and update them in this week. I will also check samples and tutorials in this week.

samples/dnn/scene_text_spotting.cpp

update opencv.bib

modules/dnn/src/model.cpp

modules/dnn/include/opencv2/dnn/dnn.hpp

- avoid 90/180 degree turns

alalek · 2020-12-01T13:10:58Z

modules/dnn/include/opencv2/dnn/dnn.hpp

+     * @return array with text detection results
+     */
+    CV_WRAP
+    std::vector<cv::RotatedRect> detect(InputArray frame) const;


Actually RotatedRect doesn't work well with perspective transformations (like this one).

Perhaps we need 4 points in API with strong order (bl, tl, tr, br - according to targetVertices) which should be used with getPerspectiveTransform() to get more accurate results.

I will try to update API on this week.

I can not open the above link. Can you share it with google drive?

It is similar to side boxes of cube from here.

However, DL models may not work with perspective transformations (EAST output doesn't know anything about that) and they just detect rotated text.

any thoughts?

It depends on different DL models. Some methods can output irregular quadrilaterals.

From my side, the perspective transformation is a temporary replacement, or just a choice for the four-point outputs. (rotated boxes and irregular quadrilaterals)

There is a popular and fast text recognition algorithm ASTER which adopts Thin Plate Spline Transformation in its rectification network. (like this).
I am not sure whether the TPS transformation is implemented in OpenCV, maybe not?
You can refer to this.

By the way, ASTER is also an algorithm from our lab, and we are glad to contribute it to OpenCV.
However, there are some things to do before.

support TPS transformation

update LSTM in modules/dnn/src/onnx/onnx_importer.cpp
we need to set these parameters non-zero, but it is not supported now.

support GRU layer

...

I suppose to work on it after this PR.

alalek

Updated TextDetectionModel API:

added quadrangle support with strong requirement for order of returned points
dropped .detectTextCountours() from TextDetectionModel_DB (replaced by quadrangles).

alalek · 2020-12-03T00:27:22Z

doc/tutorials/dnn/dnn_text_spotting/dnn_text_spotting.markdown

+Examples:
+```bash
+example_dnn_scene_text_recognition -mp=path/to/crnn_cs.onnx -i=path/to/an/image -rgb=1 -vp=/path/to/alphabet_94.txt
+example_dnn_scene_text_detection -mp=path/to/DB_TD500_resnet50.onnx -i=path/to/an/image -ih=736 -iw=736


-mp=path/to/DB_TD500_resnet50.onnx
-ih=736 -iw=736

Please check model parameters here (and above near model download links).
This set performs better: -ih=736 -iw=1280 (on IC15/test_images/img_5.jpg)

BTW, It make sense to put some defaults into TextDetectionModel_DB ctor.

Actually, DB_TD500_resnet50.onnx and DB_IC15_resnet50.onnx are prepared for different datasets (i.e. TD500 and IC15) respectively, which aims to perform better performance on each benchmark. The recommended settings do not cover the above case (use DB_TD500_resnet50.onnx on IC15 images).

If it is needed, I can train a new model with different datasets together. What is your opinion?

Thanks for explanation. That makes sense.

I can train a new model

This can be an activity after this PR merge.

Is there a tutorial on training custom models and converting existing models to onnx?

alalek · 2020-12-03T00:31:19Z

modules/dnn/include/opencv2/dnn/dnn.hpp

+     * Each result is quadrangle's 4 points in this order:
+     * - bottom-left
+     * - top-left
+     * - top-right
+     * - bottom-right


Added strong requirements for results to avoid "points reordering" in sample code.

alalek · 2020-12-03T00:32:16Z

modules/dnn/include/opencv2/dnn/dnn.hpp

+    CV_WRAP
+    void detect(
+            InputArray frame,
+            CV_OUT std::vector< std::vector<Point> >& detections,


Any thoughts about Point vs Point2f?

I think Point is okay.

modules/dnn/src/model.cpp

(1) keep points' order consistent with (bl, tl, tr, br) in unclip (2) update contourScore with boundingRect

wenqingzhang-gordon · 2020-12-03T07:23:07Z

modules/dnn/src/model.cpp

+    void setNMSThreshold(float nmsThreshold_) { nmsThreshold = nmsThreshold_; }
+    float getNMSThreshold() const { return nmsThreshold; }
+
+    // TODO: According to article EAST supports quadrangles output: https://arxiv.org/pdf/1704.03155.pdf


You can find that the performance of QUAD is worse than RBOX in Table. 3 and the authors do not provide official code and models.
Some good re-implementations of EAST only support RBOX.
TF: https://github.com/argman/EAST
PyTorch: https://github.com/songdejia/EAST

alalek

LGTM 👍

@HannibalAPE @bhack Thank you for contribution!

wenqingzhang-gordon · 2020-12-04T02:18:26Z

Thank you. I really appreciate your help. @alalek @vpisarev @dkurt @bhack @berak

[GSoC] High Level API and Samples for Scene Text Detection and Recognition * APIs and samples for scene text detection and recognition * update APIs and tutorial for Text Detection and Recognition * API updates: (1) put decodeType into struct Voc (2) optimize the post-processing of DB * sample update: (1) add transformation into scene_text_spotting.cpp (2) modify text_detection.cpp with API update * update tutorial * simplify text recognition API update tutorial * update impl usage in recognize() and detect() * dnn: refactoring public API of TextRecognitionModel/TextDetectionModel * update provided models update opencv.bib * dnn: adjust text rectangle angle * remove points ordering operation in model.cpp * update gts of DB test in test_model.cpp * dnn: ensure to keep text rectangle angle - avoid 90/180 degree turns * dnn(text): use quadrangle result in TextDetectionModel API * dnn: update Text Detection API (1) keep points' order consistent with (bl, tl, tr, br) in unclip (2) update contourScore with boundingRect

berak reviewed Jun 17, 2020

View reviewed changes

asmorkalov added category: samples GSoC labels Jun 18, 2020

vpisarev changed the title ~~add scene text detection and recognition samples~~ [GSoC] add scene text detection and recognition samples Jun 24, 2020

wenqingzhang-gordon changed the title ~~[GSoC] add scene text detection and recognition samples~~ [GSoC] High Level API and Samples for Scene Text Detection and Recognition Jul 7, 2020

wenqingzhang-gordon mentioned this pull request Jul 7, 2020

[GSoC] test data for scene text detection and recognition (High Level API) opencv/opencv_extra#773

Merged

bhack reviewed Jul 13, 2020

View reviewed changes

samples/dnn/scene_text_detection.cpp Show resolved Hide resolved

wenqingzhang-gordon force-pushed the text_det_recog_demo branch from 1b4d47d to 9c81421 Compare July 16, 2020 14:30

wenqingzhang-gordon force-pushed the text_det_recog_demo branch from 6fbaf85 to cb2279e Compare August 3, 2020 09:21

wenqingzhang-gordon force-pushed the text_det_recog_demo branch from 8ba1956 to 01ebd94 Compare August 18, 2020 14:02

APIs and samples for scene text detection and recognition

7ead0f6

wenqingzhang-gordon force-pushed the text_det_recog_demo branch from 0daa5de to 7ead0f6 Compare August 18, 2020 14:29

Merge branch 'master' into text_det_recog_demo

d1a2df9

alalek reviewed Sep 3, 2020

View reviewed changes

wenqingzhang-gordon requested a review from alalek September 5, 2020 03:26

alalek reviewed Sep 5, 2020

View reviewed changes

wenqingzhang-gordon force-pushed the text_det_recog_demo branch from e633676 to 8a12190 Compare September 7, 2020 05:40

dkurt reviewed Sep 7, 2020

View reviewed changes

alalek reviewed Nov 19, 2020

View reviewed changes

wenqingzhang-gordon commented Nov 20, 2020

View reviewed changes

dnn: refactoring public API of TextRecognitionModel/TextDetectionModel

f59aa6d

alalek reviewed Nov 21, 2020

View reviewed changes

This was referenced Nov 22, 2020

Add option for NMS for boxes with different labels #18891

Merged

objc: fix handling of std::vector<std::vector<T>> return type #18905

Merged

wenqingzhang-gordon commented Nov 26, 2020

View reviewed changes

samples/dnn/scene_text_spotting.cpp Outdated Show resolved Hide resolved

update provided models

63672da

update opencv.bib

wenqingzhang-gordon mentioned this pull request Nov 26, 2020

DNN module: ONNX importer can not handle models generated by Pytorch==1.7.0 #18934

Open

4 tasks

dnn: adjust text rectangle angle

ca1f5d9

alalek reviewed Nov 27, 2020

View reviewed changes

modules/dnn/src/model.cpp Outdated Show resolved Hide resolved

remove points ordering operation in model.cpp

9aebabe

wenqingzhang-gordon force-pushed the text_det_recog_demo branch from 0f10240 to 9aebabe Compare November 30, 2020 09:19

dkurt reviewed Nov 30, 2020

View reviewed changes

modules/dnn/include/opencv2/dnn/dnn.hpp Show resolved Hide resolved

wenqingzhang-gordon and others added 2 commits November 30, 2020 22:58

update gts of DB test in test_model.cpp

b4a0300

dnn: ensure to keep text rectangle angle

e403b7b

- avoid 90/180 degree turns

alalek reviewed Dec 1, 2020

View reviewed changes

dnn(text): use quadrangle result in TextDetectionModel API

b705253

alalek reviewed Dec 3, 2020

View reviewed changes

dnn: update Text Detection API

1eb24da

(1) keep points' order consistent with (bl, tl, tr, br) in unclip (2) update contourScore with boundingRect

wenqingzhang-gordon commented Dec 3, 2020

View reviewed changes

alalek approved these changes Dec 3, 2020

View reviewed changes

alalek merged commit 22d64ae into opencv:master Dec 3, 2020

This was referenced Dec 5, 2020

samples: replace regex #19018

Merged

dnn(test): adjust EAST test tolerance #19074

Merged

alalek mentioned this pull request Apr 9, 2021

(5.x) Merge 4.x #19885

Merged

Uh oh!

Conversation

wenqingzhang-gordon commented Jun 17, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Pull Request Readiness Checklist

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

berak Jun 17, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

bhack commented Jul 13, 2020

Uh oh!

Uh oh!

vpisarev commented Jul 27, 2020

Uh oh!

wenqingzhang-gordon commented Jul 27, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

wenqingzhang-gordon commented Jul 28, 2020

Uh oh!

vpisarev commented Jul 29, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

wenqingzhang-gordon commented Jul 30, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

alalek left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

alalek commented Sep 5, 2020

Uh oh!

alalek commented Sep 5, 2020

Uh oh!

alalek left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

alalek commented Sep 5, 2020

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

wenqingzhang-gordon commented Jun 17, 2020 •

edited

Loading

berak Jun 17, 2020 •

edited

Loading

wenqingzhang-gordon commented Jul 27, 2020 •

edited

Loading

vpisarev commented Jul 29, 2020 •

edited

Loading

wenqingzhang-gordon commented Jul 30, 2020 •

edited

Loading

wenqingzhang-gordon commented Nov 19, 2020 •

edited

Loading

alalek left a comment •

edited

Loading

wenqingzhang-gordon Nov 20, 2020 •

edited

Loading