Playing the game with Azure Cognitive Services
summary
Demo shows a working pipeline: custom vision trains a card-recognizer with ~50 images (quick train 5–10 minutes, iteration 8) and returns tags with probabilities (examples 18.6% and 87%), called from a low-code app built with Power Apps + Power Automate plus Azure Blob and an Azure Function that picks the top prediction. A separate proof uses form recognizer trained on ~5 receipts in ~30 minutes to extract totals (example €10.50). Cognitive containers run locally for lower latency and privacy but require periodic internet check-ins.
AI-written summary · about
chapters (14)
- 0:00Introduction and welcome
- 1:51Why use AI and cognitive services
- 4:01Advantages of Azure Cognitive Services
- 6:07Machine learning process overview
- 10:02Overview of cognitive services categories
- 13:59Vision services deep dive
- 18:37Custom Vision service explained
- 27:45Demo: Playing card recognition
- 37:22Demo: Blackjack card value calculator
- 40:14Demo recap and Form Recognizer intro
- 41:17Form Recognizer service explained
- 44:33Demo: Parking receipt extraction
- 49:01Cognitive containers overview
- 54:00Resources and closing remarks
related talks
57mEasier AI allows everyone to build smart apps
01h 03mEasier AI allows everyone to build smart apps
42mProduct Recognition in Retail/E-Commerce with Azure AI
01h 04mBuilding and Training your own Custom Image Recognition AI
42mProduct Recognition in Retail/E-Commerce with Azure AI
50mAzure AI, Power new possibilities for every organization
46mBuilding accessible experiences with AI
49mMake your applications intelligent (...)