VideoTextReader is a Google Chrome extension developed for the purpose of explaining text present in frames and videos chosen by the user. The functionality of this extension is based on GPT-3.5 and Amazon Textract.
This product was mainly inspired by one of GPT-4's newest feature, the ability of analyzing and interpreting images and create text, like descriptions, based on them.
Using the terminal, clone this repository using the following command:
$ git clone https://github.com/oscardcamargo/VideoTextReader.git
Open Google Chrome and go to chrome://extensions/. Activate Developer Mode in the top right. Click Load Package and select the repository.
You can now try our extension to create short descriptions of any frame with text in any YouTube video.