- Use a webcam as a mirror to view your data identities
- Flask app that accepts basic user information to display your Facebook and New York City data over a live OpenCV facial detection visualization
- An exploration tool that allows you to explore your own data, as well as aggregated data of other New York residents and crowdsourced FB ads dataset from Propublica (https://www.propublica.org/datastore/dataset/facebook-ad-categories)
- Developed for the 2020 Data x Design Exhibit
├── README.md
├── main_fb.py < - flask app to run fb web interface >
├── main_gov.py < - flask app to run gov web interface >
├── ageGenderDetect.py < - helper functions to detect face/age/gender >
├── facebookStyling.py < - helper functions for styling facebook video stream >
├── govStyling.py < - helper functions for styling gov video stream >
├── models <- ML prototxt and caffe models for face/age/gender detection >
├── static <- CSS files >
│ ├── styles
│ ├── main.css
├── templates <- HTML files >
│ ├── index_fb.html < - display gov form UI >
│ ├── index_gov.html < - display facebook form UI >
│ ├── video.html < - display video stream UI >
├── portraits <- Placeholder for data portraits >
├── data <- datasets used for app >
│ ├── fb_data <- Folder that contains ProPublica FB Ads data or uploaded personal fb ads data >
│ ├── ├── facebook <- Placeholder for personal dataset>
│ ├── 311.json <- Dataset of 311 complaint data, by zipcode >
│ ├── income.json <- Dataset of average income data, by zipcode, county >
│ ├── zipmap.json <- Dataset that maps zipcode to county >
│ ├── fb_propublica.json <- Propublica FB ads dataset >
*Note: in the repo models, portraits, and data folders are in root.zip file
Requirements:
- Python3
- Computer with Webcam
pip install requirements.txt
To get started (while the flask app is in development), run:
python main_fb.py --ip 0.0.0.0 --port 8080
or
python main_gov.py --ip 0.0.0.0 --port 8080
