Analyticat was built during the StirHack 2017 MLH event, over the course of 24 hours.

• Register IoT video sources to gain analytics about your customers; including:
  • a measure of recurring customers on an individual, but anonymous level, using AWS-powered facial recognition (you will not be able to track individual customers)
  • estimate demographics including gender, age etc.
• View analytics in a clean dashboard. You can view analytics over various discrete timesteps (hours, days, weeks, etc.) for your registered video sources (or groups thereof).
• Utilise our own REST API so that you can use your own analytics data for something more unique!

• The analyticat.net frontend mockup is available to view.
• Our REST API is somewhere in the region of 50% finished:
  • finished:
    • emotional and sentiment analysis of people in uploaded images
    • check similarity between two people in uploaded images
  • not finished:
    • demographic analysis (age, gender, etc.)
    • coordination and interfacing between the IoT layer, the front end and our AWS MySQL database
• We currently have a [camera watcher](https://github.com/Leah Hirst/StirHack/blob/master/preproc/watcher.py) which continually watches over the default camera - whenever a face is detected the image is saved (but would be uploaded via our REST API upon completion).