Science Art in the Cloud: How I built a Museum Exhibit
summary
Builds an interactive museum kiosk that photos passersby, aligns facial landmarks, and averages the last ~10 images into a blended portrait; it runs on a UWP client (Raspberry Pi or PC) that calls a Python Azure Functions backend, uses Face API for landmarks, OpenCV for affine transforms, blob storage to fetch the most recent images, and numpy average to blend them. The exhibit deploys to multiple sites and also exposes a web/chatbot front end at the provided URL, and it works live.
AI-written summary · about
chapters (11)
- 0:00Introduction to the talk
- 1:05Speaker background and motivation
- 2:07Concept of people blending explained
- 5:30Cognitive portrait demo and tools
- 12:17Building the museum exhibit
- 15:48UWP app for face detection
- 18:16Azure Function for image blending
- 25:02Deploying to Azure cloud
- 29:20Virtual exhibit and chatbot
- 37:13Technologies recap and resources
- 38:15Call to action: blend your photo
related talks
31mNon-traditional usage of Cloud for Science Art
50mScience Art in the Cloud: How Microsoft Azure helped me to Express Myself
32mSelf-Portraits in JavaScript
39mAI + JavaScript Rocks
38mHow Artificial Intelligence has changed your life
48mThe Future of Machine Learning & JavaScript
33mSee More, Be More
21mBring AI-Based Search to Your Web App