022-10-23 23:13:54 | DEBUG - Downloading objects.yaml from https://raw.githubusercontent.com/QIN2DIM/hcaptcha-challenger/main/src/objects.yaml 2022-10-23 23:14:02 | WARNING - None 2022-10-23 23:14:06 | WARNING - None 2022-10-23 23:14:10 | WARNING - None 2022-10-23 23:14:14 | WARNING - None 2022-10-23 23:14:17 | WARNING - None 2022-10-23 23:14:21 | WARNING - None 2022-10-23 23:14:25 | WARNING - None 2022-10-23 23:14:29 | WARNING - None

this issue persist

The object in this case is cup of coffee, it takes screenshot, then take again and so on

It doesnt work when changing chrome language to english and changing get_labels to return label in english But the image classifier just takes chinese

Introduction

Usage

You can run the demo project associated with the solution by running the following command:

# hcaptcha-challenge/src
python3 main.py demo_v3

Linked pull requests

@beiyuouo https://github.com/QIN2DIM/hcaptcha-challenger/commit/30ae3915b95e65cedfd9d0d8a9896bfdd9cd114f

Demo

Hi,

First i wanted to thank you for your work !

I just downloaded your project, i installed the required dependencies and i launched the program using the same line as your readme gif. The problem is, it doesn't succeed resolving the hcaptcha. How come ? Is there something i'm missing ?

Hello, I have no error but the script crashes, I have not modified the source code but just added to my private project. The captcha solver is called at the end to solve a captcha.

Here is the error: C:\Users\user\Desktop\discord>python account.py

DevTools listening on ws://127.0.0.1:1089/devtools/browser/70f17e73-3f59-4ffb-a093-195fd79dfa61 [19060:15084:0831/182935.625:ERROR:device_event_log_impl.cc(214)] [18:29:35.626] Bluetooth: bluetooth_adapter_winrt.cc:1074 Getting default adapter failed. 2022-08-31 18:29:45 | DEBUG - >> Challenge [ArmorCaptcha] Handle hCaptcha checkbox 2022-08-31 18:29:47 | DEBUG - >> Challenge [ArmorCaptcha] Get label - label=「Please click on each image containing a lion with closed eyes.」 2022-08-31 18:29:47 | DEBUG - >> Challenge [ArmorCaptcha] Download challenge images - timeit=0.2s 2022-08-31 18:29:48 | WARNING - >> ALERT [ArmorCaptcha] Types of challenges not yet scheduled - label=「Please click on each image containing a lion with closed eyes.」 prompt=「Please click on each image containing a lion with closed eyes.」 screenshot=C:\Usersuser\Desktop\discord\database\temp_cache\captcha_schot\1661963387. Please click on each image containing a lion with closed eyes..png site_link=https://discord.com/register issues=https://github.com/QIN2DIM/hcaptcha-challenger/issues?q=Veuillez%20cliquer%20sur%20chaque%20image%20contenant%20un%20lion%20aux%20yeux%20ferm%C3%A9s.

C:\Users\user\Desktop\discord>

I enclose an extra picture.

Note: This is an error i got which is from my implementation of the ai.

  File "C:\Bot\dislock\main.py", line 411, in captcha_solver
    self.challenger = ArmorCaptcha(dir_workspace=DIR_CHALLENGE, dir_model=DIR_MODEL, lang='en', debug=True,
  File "C:\Bot\dislock\hcaptcha_challenger\core.py", line 126, in __init__
    self.pluggable_onnx_models = self.pom_handler.overload(
  File "C:\Bot\dislock\hcaptcha_challenger\solutions\resnet.py", line 203, in overload
    return {
  File "C:\Bot\dislock\hcaptcha_challenger\solutions\resnet.py", line 204, in <dictcomp>
    finger: FingersOfTheGolderOrder(finger, dir_model, path_rainbow)
  File "C:\Bot\dislock\hcaptcha_challenger\solutions\resnet.py", line 93, in __init__
    super().__init__(onnx_prefix, f"{onnx_prefix}(ResNet)_model", dir_model, path_rainbow)
  File "C:\Bot\dislock\hcaptcha_challenger\solutions\resnet.py", line 41, in __init__
    self.net = cv2.dnn.readNetFromONNX(self.onnx_model["path"])
cv2.error: OpenCV(4.5.5) D:\a\opencv-python\opencv-python\opencv\modules\dnn\src\onnx\onnx_importer.cpp:709: error: (-215:Assertion failed) model_proto.has_graph() in function 'cv::dnn::dnn4_v20211220::ONNXImporter::populateNet'

Source: https://github.com/Vinyzu/DiscordGenerator/blob/main/main.py#L411

Hi, I respect your work and know how much it is useful, but I can't integrate it with my python code I want you to give a standard onnx format model so that everyone like me can make prediction and use it anyway wanted maybe for even classification task Does all predictions made by yolov6? And why don't you use yolov7 as it is fast Can you give links about guide on how did you actually trained the model for custom classification like dog_with_open_eyes I also want to use it to solve recaptcha Basically, I am making a pypi version of library that can handle every captcha with single line like this

solver = CaptchaSolver().setCaptchaTypeAs#anylikehcaptcha.solve()

I actually got success with very good accuracy using efficientnet before but it can't predict new classes that hcaptcha made recently Also, I have solved too many classes of recaptcha, text captcha, upside down captcha, duplicate image finder captcha, slide captcha, and many more, generally 9 types supported by it I want you to contribute in it by giving me and teaching me how to do custom classification task, dataset length required for it and give me yolov6 in h5 or onnx format Thanks

If i know the site key can i get many pictures to train? I have tried many times but it is very difficult to get many pictures, at most 80 ~ 100 pictures per 1000 times. Is there any way to get the exact picture i need?

In extreme tests, the effect of the following model does not meet the expectations, and we plan to release them in the future.

• [ ] living_room
• [x] bedroom
• [ ] smiling_dog
• [x] conference_room

We will release the heterogeneous scaffolding in v1.0 and upload the project package to pypi platform, then players can use pip to install and use hcaptcha-challenger.

But before that, we will absorb as many suggestions from players as possible, and we will also re-standardize the model of the scaffolding interface so that it can be called in various suitable occasions.

Prompt[en]

Please click each image containing a rabbit

New types of challenge

New prompt (for ex. Please select all the 45th President of the US)

Sitekey

https://accounts.hcaptcha.com/demo?sitekey=edc4ce89-8903-4906-80b1-7440ad9a69c8

Sitelink

https://accounts.hcaptcha.com/demo?sitekey=edc4ce89-8903-4906-80b1-7440ad9a69c8

Screenshot of the challenge

Prompt[en]

red roses in a garden

New types of challenge

New prompt (for ex. Please select all the 45th President of the US)

Sitekey

f5561ba9-8f1e-40ca-9b5b-a0b3f719ef34

Sitelink

https://accounts.hcaptcha.com/demo?sitekey=f5561ba9-8f1e-40ca-9b5b-a0b3f719ef34

Screenshot of the challenge

PS C:\Users\User\Desktop\Temp\hcaptcha-challenger-main 213821938123\hcaptcha-challenger-main> python main.py test
Traceback (most recent call last):
  File "C:\Users\User\Desktop\Temp\hcaptcha-challenger-main 213821938123\hcaptcha-challenger-main\main.py", line 10, in <module>
    from examples import demo_selenium, demo_challenge, demo_install, demo_classify
ImportError: cannot import name 'demo_selenium' from 'examples' (C:\Users\User\AppData\Local\Programs\Python\Python310\lib\site-packages\examples\__init__.py)

Prompt[en]

Are all red dots on the longest black line?

New types of challenge

New approach (for ex. image watermark)

Sitekey

adafb813-8b5c-473f-9de3-485b4ad5aa09

Sitelink

https://accounts.hcaptcha.com/demo?sitekey=adafb813-8b5c-473f-9de3-485b4ad5aa09

Screenshot of the challenge

    @QIN2DIM Can you make another repo for the Pypi Release/ Make an Update

Basicly, these are my two issues:

1. It has requirements such as undetected-chromdriver that it doesnt use
2. The Directory where the Models are saved should be fixxed (for example in the local pip dir) as if u use hcaptcha_chalenger in two projects you download the models a second time.

Originally posted by @Vinyzu in https://github.com/QIN2DIM/hcaptcha-challenger/issues/226#issuecomment-1295892623

This is a list of hCaptcha Sitekeys i found all arround the internet:

3b35d792-d035-4ea9-ad4e-61179582a5ee
3ceb8624-1970-4e6b-91d5-70317b70b651
28982ab4-aeef-4200-b9cc-28af1e23e377
ad252538-16a2-4618-a23c-51485c2c9622
a5f74b19-9e45-40e0-b45d-47ff91b7a6c2
28432475-d840-43fc-8b44-ccbf78ce90c2
ace50dd0-0d68-44ff-931a-63b670c7eed7
f5561ba9-8f1e-40ca-9b5b-a0b3f719ef34
91e4137f-95af-4bc9-97af-cdcedce21c8c
adafb813-8b5c-473f-9de3-485b4ad5aa09
c86d730b-300a-444c-a8c5-5312e7a93628
edc4ce89-8903-4906-80b1-7440ad9a69c8
9a5da798-adac-4c50-9c7e-79d9f92bd4c0
4c672d35-0701-42b2-88c3-78380b0db560
13257c82-e129-4f09-a733-2a7cb3102832
f1592dee-67b6-4670-9062-649933011aa2
a9b5fb07-92ff-493f-86fe-352a2803b3d0
8adc2f3e-e343-4ef9-bf6c-275a8a2818ab
50d5f7a9-8aed-4244-aea5-20a291951e93
c1ec07d4-a6a8-4e42-b8c7-f0de89453007
b364b1fd-e3d8-4d24-8c41-77a19604b00d
7734ec9b-f8cb-44b2-9fac-3a502cb4f1bf

Hope this helps someone! :D