Let's Build a Doppelganger Game!
summary
A working doppelganger app gets built using Azure Custom Vision, an Angular front end and Node calling a Bing Image Search API to collect training images, with serverless Azure Functions automating upload and training. The deployed prediction endpoint returns similarity scores (examples: 61% vs 95% or a 60% Arnold / 38% Peter split), accuracy depends on dataset size and quality (recommend at least 50 images per class, realistically thousands), and the demo works but the model is noisy.
AI-written summary · about
chapters (6)
- 0:00Intro
- 0:49What are doppelgangers
- 8:39Our goal
- 8:57The process
- 11:07Creating a new project
- 17:36Using Angular
related talks
23mBuild a look alike engine with machine learning and Angular
01h 04mBuilding and Training your own Custom Image Recognition AI
01h 01mGetting Started with Computer Vision
44mHop on the serverless adventure
01h 03mEasier AI allows everyone to build smart apps
42mProduct Recognition in Retail/E-Commerce with Azure AI
46mCreate Smart Angular Apps Using Azure Computer Vision
58mPlaying the game with Azure Cognitive Services