Claritag is a web app that categorises photos in a user’s Dropbox folder. The categorisation is done through the Clarifai API, which provides an endpoint for us to send through a link to the image. Fortunately Dropbox provides the
.media API call, which allows us to get a public link to the file so we can plug it into Clarifai’s API. If we did not have this available, we would need to provide some public facing server, such as AWS S3 or DigitalOcean in which to save these images temporarily for analysis.
We list all top-level JPEGs and PNGs inside our
/Photos folder in Dropbox, and utilise the above API call to collect the tags from Clarifai’s API. We store the top 10 tags in a database entry (with a horribly un-normalised schema; sorry, this is hackathon code!) for later collection.
When a user browses to
http://localhost:5000, they will be presented with the list of all tags found in the database, linking to a separate page to list all images which have been found for that image.
This project was built using the following technologies:
- Flask: Flask is a microwebframework built in Python, with a focus on simplicity and ease of use.
- Dropbox: Dropbox is a file hosting service with a _host of native applications_
- Clarifai: An image and video recognition API.
- MySQL: An open-source relational database management system.
Find the project on GitLab .