Skip to content

Latest commit

 

History

History
48 lines (26 loc) · 2.67 KB

initial-funding-concept.md

File metadata and controls

48 lines (26 loc) · 2.67 KB

Project Title

Stealing Ur Feelings



Concept Summary

In 140: Stealing Ur Feelings is an interactive film that reveals how Snapchat can use your face to secretly collect data about your emotions

Everyone loves those Snapchat-style AR filters that give you dog ears or whiskers. These apps, powered by machine learning and AI, have exploded in popularity - with Facebook AR arriving on Instagram, Apple's Animoji tech and more.

But while you're using them, you could be leaking some very private data.

Using AI and machine learning, AR filter apps can perform an operation known as emotion recognition... and literally steal your feelings.

Stealing Ur Feelings will be a short web-based interactive documentary film that reveals the wild science of emotion recognition.

We'll access the user's webcam and integrate a live AR filter into the experience. Instead of turning the user into a dog, we'll create unexpected filters that punctuate our story.

We'll take the user into The Matrix-like computer world of feature detection, displaying the wireframe of their face as they watch.

And we'll reveal how algorithms can determine if they're happy or sad, and how corporations can correlate their feelings with the content they consume to do some Not Nice Things.

I've got a demo I can show you!



Suitability

Brett Gaylor selected my previous interactive film, Weird Box, for exhibition at Mozilla Festival 2017. Weird Box explored similar privacy issues relating to our photo/video media and the social apps we use to share it.

As a software engineer, I've done original research in the field of machine learning-based face classification: At Recurse Center, I researched the potential of the WebAssembly MVP for broadly accessible machine learning-based CV on the web. As part of my research, I wrote the first wasm implementation of the Viola-Jones research paper on rapid face detection. As someone who has debugged machine learning algorithms, I'm intimately acquainted with the social biases that can be introduced; to my horror, I have at times accidentally introduced these biases myself.

My work has received many accolades for its depictions of online life. I won a PromaxBDA Gold Award for collaborating with Key & Peele on their Social Influencers digital sketches. Before that, I won a Vimeo Staff Pick and received international press for Chillr, an internet project about an imaginary mobile app.

https://bit.ly/2tdvs4O

https://bit.ly/2yyINIw

https://bit.ly/2lt7yMI

https://bit.ly/2tqRnTK

https://bit.ly/2yucb2y



On which Internet Health Issue(s) does your concept focus?

Privacy & Security