Inspiration
Like every other programmer, I use youtube to watch coding tutorials. The process of learning is definitely fun but trying to follow a tutorial can be hectic. Shifting between your browser and IDE and typing the code can waste precious time. I wanted to counter this everyday issue and help the CS community!
What it does
YType enables the user to select and take a screenshot of an image or video that contains a piece of code. Then it converts the code to text so that you can copy and paste those lines to your own code.
How I built it
I am using a Google extension called Nimbus to select an area of the browser to take a screenshot. That screenshot is saved into the directory folder for my python code. When the code is run it sends that image to Google Vision through the dense document text detection feature that returns the text version.
Challenges I ran into
I found it difficult to find datasets to train the vision model.
Accomplishments that I'm proud of
Completed my first solo hackathon project.
What I learned
I learned the use of Google vision for OCR and image classification.
What's next for YType
I want to create a Google Chrome extension that does all of the processes so the user can get the text from a video without ever leaving the webpage.
Built With
- google-vision
- nimbus
- python
Log in or sign up for Devpost to join the conversation.