I just finished my first ever blog post. It’s part 1 in a mini-series about captcha recognition.
It kind sets the stage by introducing the dataset and using the multilabel-classification approach from lesson 3 to give a proof of concept that CNN’s are definitely capable of captcha recognition.
The underlying notebook can be found here, the dataset here and the blogpost right here.
This is my first time writing a blog post, so I’d appreciate any input. Thanks!