WWW Filter | Devpost

The Beautiful Extension Itself
NO NO IMAGE Filtered and Defeated

Why is it important?

For parents and kids, the internet can be a scary place. There’s lots of bad language, inappropriate imagery, viruses, etc. The internet can be so much fun for kids, not to mention a useful learning tool, but the risk of them seeing something that could be bad for them can be worrying for parents. WWW Filter gives parents and guardians confidence that their kids can surf the web in peace, without having to ban it all together.

What does it do?

After downloading the free extension, a parent or guardian clicks on the extension to turn it on to start their kid’s safe browsing sessions. While in safe mode, this extension blocks credit card payments and filters out NSFW content. Websites on the no-fly list are rerouted. Inappropriate language on websites is censored, and images determined inappropriate by Computer Vision are blurred. Anytime the parent or guardian wants to turn it off safe browsing, all they have to do is enter their password!

asdfmovie

How did we build it?

Frontend and backend are both JavaScript, with Node.js, HTML, and CSS. Image recognition is done through Microsoft Azure’s Computer Vision API, and text replacement is done with Regex. Test page here

What challenges did we run into?

At about 3am, the team had to abandon the initial project idea of AR map directions for inside buildings. There was difficulty setting up the environment for the AR to run and the GPS recordings varied by more than 40m when inside, making the project infeasible :'( .

thumbs up kid

What’s next for WWW Filter?

There’s more potential for customization in allowing parents to choose what they will allow their kids to see. There’s also potentially unsafe video and audio on the web, so being able to detect profanity and NSFW content from those media forms as well would help contribute to WWW Filter’s goals. It would also be able to support multiple devices, and possibly give the user more preferences.

Frontend Github Repo HERE

The extension will be available on the Chrome Web Store once approved!!!

Built With

cheerio
computer-vision
css
html5
javascript
node.js
regex

Submitted to

TAMUhack 2021

Created by

I implemented regex for text replacement to censor inappropriate language and I contributed to the devpost submission project details. In the original project idea (before it was scrapped) I was in charge of the backend. I constructed graphs from geospacial coordinates and implemented the Dijkstra search algorithm to route clients between the nearest start node to their current location and a destination node of their choosing. The issue was that the client side GPS coordinates where too inaccurate to properly tell where the person was when inside, hence the change in projects.

Andrew Teske
I came up with the idea for the project, wrote and formatted the markdown, learned video editing, and helped out with decision making/researching

Nickzad Rafieha
I worked on the front-end of the extension and implementing the Computer Vision API in our back-end. On our original idea, I worked on getting AR to work with React Native, creating our custom location data collection software, and a bunch of geolocation math.

Rishabh Tatia
I'm a passionate developer who loves building products that make life a little easier!
Made the presentation, the UI, the graphics, and contributed to the design.

Michael Stewart

Updates

Andrew Teske started this project — Jan 31, 2021 01:27 PM EST

Leave feedback in the comments!

Log in or sign up for Devpost to join the conversation.