Published in


Making the Web More Accessible With AI

Person reading Braille stock image (src)
Demo of the chrome extension in action

Why I made Auto Alt Text:

Email Text: “Hi Abhinav, I found your flask-base project and think it is definitely going to be a great fit for my next project. Thanks for working on it. I also just wanted to let you know that you should put some alt descriptions on your readme images. I’m legally blind and had a tough time making out what was in them :/ From REDACTED”

Artificial Intelligence to the Rescue:

  1. They aren’t responsive and take a long time to return a caption.
  2. They are semi-automated (i.e. relying on humans to manually caption images on demand).
  3. They are expensive to create and maintain.
Examples of im2txt in action from the im2txt Github Repository

Technical Details of im2txt:

Overview of the structure of the neural network (from the im2txt Github repository)

Problems out of the box + Lamdba:

AWS API Gateway + AWS = heart (src)
Image with likely probabilities of what is in the image according to the API
Auto Alt Text Chrome Extension in action


Categories of images in the COCO dataset



  • Follow me: I primarily publish what I am doing on my Medium. If you like this post, I’d appreciate a follow :) In the coming months I’ll be publishing a couple more “how to” guides for using AI/tensorflow to solve real world problems. I’ll also be posting some JS explainers/tutorials in the near future.
  • Link to the Chrome Extension download page
  • Link to the Auto Alt Text Lambda API Github repository



Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Abhinav Suri

UPenn 2019. Computational Biology. Co-Director @Hack4Impact, Intro to JS Lecturer. Writer for @freeCodeCamp, @hackernoon, @TDataScience, and @JavaScriptDaily