We wanted to enable safe browsing for people for whom browsing the web is inherently risky. This includes everyone from recovering drug addicts, to PTSD victims, to children.

What it does

Distill is a Chrome extension that censors photos and words based on the machine-learning API Clarifai and different genres of censorship.

How we built it

We used JavaScript, Clarifai API, CSS, and HTML. Using such a wide variety of programming skills meant we had to work as a team to balance each other's strengths and weaknesses.

Challenges we ran into

Working with the number of calls to the Clarifai API became problematic, we compromised by utilizing small-batch calls to strike a balance between image processing speed and browser responsiveness. We also added pleasant blur and ease effects to make the small amount of added latency more pleasing for the end user.

Accomplishments that we're proud of

Our use of CSS to create an aesthetically pleasing user experience on top of a very powerful image analytics API gave us a surprisingly polished yet powerful app in a very short amount of time.

What we learned

We learned that APIs can be extremely powerful and also experimented with more web-centered development.

What's next for Distill

We would really like to implement user-generated censor genres for more specific censoring and user-tagged photos to train our API for more confident results when dealing with subjects our users don't want to see.

Share this project: