Skip to content
/ altify Public
forked from ParhamP/altify

Uses deep learning to caption images in an HTML file and fills out its alternative text attributes with the related caption

License

Notifications You must be signed in to change notification settings

nndnha/altify

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

43 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Altify

Image for demo

Altify automizes the task of inserting alternative text attributes for image tags. Altify uses Microsoft Computer Vision API's deep learning algorithms to caption images in an HTML file and returns a new HTML file in which alt attributes are filled out with their corresponding captions.

Notice: Altify will now ignore any image tag whose alt attribute has content or is just an empty string. (In compliance with standard web practices)

Dependencies

  • BeautifulSoup
  • Python 2.7
  • html5lib

Install and Usage (Latest Version)

1) Get a Microsoft API Key for Free

https://www.microsoft.com/cognitive-services/en-us/sign-up.

2) Install via pip

Open up terminal and enter: pip install altify

3) Use

altify path_to_your_html api_key

4) Enjoy!

A new HTML file called altify.html is created next to the HTML file you selected. (Or Desktop, when using older versions)

How It was Built

  1. Parses the html using BeautifulSoup.
  2. Find all the image tags.
  3. Stream images using uploads.im API, and then with its URL, send request to Microsoft's API to caption.
  4. Filter images that are smaller than 200px width.
  5. Fill out the alt attributes for all the images.
  6. Write an edited HTML file next to the file you selected.

Captioned Images Samples

Image for demo

Donald Trump wearing a suit and tie

Image for demo

A piano keyboard

Image for demo

A squirrel eating

Image for demo

A close up of a cat looking at the camera

Image for demo

A woman wearing a red hat

Image for demo

A small boat in a lake surrounded by mountains

Disclaimer

Humans are currently better at captioning images than machines. Use responsibly!

About

Uses deep learning to caption images in an HTML file and fills out its alternative text attributes with the related caption

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Python 100.0%