VKCaptchaBreaker_v1.6.2

Changelog:

• Fixed a bug when base64 image was downloaded over and over, which on some devices may cause image "blinking"
• Added support to a new captcha form
• Changed icon
• Added restriction to some unwelcome page URLs on which captcha bypass is not supported

VKCaptchaBreaker_v1.6

v1.6 Improvements:

• Images with captcha are now downloaded in background script which makes it possible to bypass CORS restrictions;
• Models are baked into background script, no more unnecessary binary files;
• Communication between page and background script is now carried out via base64 encoding, which takes less time compared to hex encoding;
• Page script's logic is now hidden with encapsulated anonymous function;
• New model, 30% relative improvement in recognition time.

VKCaptchaBreaker_v1.5

v1.5 Improvements:

• Recognition logic has migrated to the background script, causing faster recognition, lightweight content script and better extension experience in general;
• ONNX tensor's re-usage, so memory will be utilized efficiently, and V8's Garbage Collector will less interrupt recognition process;
• ONNX's trick, when we use float32 tensor from the first model as int32 tensor, simply changing its type, meaning no actual types casting is required;
• Fixed bug in Vkontakte's login form setup, when the extension has been pressing a login button in an infinite loop;
• The extension now works at backend script on Android devices (tested in Kiwi browser);
• Updated tensor preprocessing (before sending array to the backend script filter alpha-channel);
• Updated model weights (previous versions actually used 850k trained images (~82% accuracy). Now latest available model, trained on 1500k images with knowledge distillation is used (~91% accuracy).

VKCaptchaBreaker_v1.4

v1.4 Changelog:

Extension is optimized via JS MutationObserver utilization on document.body
That implies:

• More stable behavior, so listeners won't occasionally break if the node is not presented on the webpage;
• Changed slow document.query to much faster document.getElementsByClassName and document.getElementsByName;
• Eliminate slow forEach loop in observer callback, instead fast DOM-selectors are used for changes monitoring;
• Timers eliminated from recognition function and observer callback.

Introduced multitab support:

• From now on switching to a different tab still will listen to a captcha form;

Handling different captcha scenarios:

• As a side-product version 1.4 added captcha bypass support for 'like', 'add to friends', and other actions, which may require captcha bypass.

Known issue of version 1.4:
If the captcha is required during login, the extension can't bypass it due to different symbols length (6-7 compared to standard 4-5)

VKCaptchaBreaker_v1.3

From now on the extension only works with the mobile version of the site (m.vk.com)

v1.3 Changelog:

• Migrated to m.vk.com domain due to the fact that Vkontakte change their captcha to ReCaptcha (on mobile version there are still old ones);
• Fixed a bug when the extension is working and recognition models aren't loaded after updating the page;
• Fixed a bug when the extension is working and automatic recognition crashes on non-messages pages without captcha field;

VKCaptchaBreaker

Chrome Extension added
HTML demo added