Candy Crush QA AI Saga
summary
A playable AI bot takes screenshots, detects UI elements with image hashing and OpenCV, clicks coordinates through an automation framework, and logs a graph of 100s of screens; accuracy improves by synthetically generated images and training on other casual games. Text uses Google Vision API and TF‑IDF style similarity, audio testing is planned, and debug commands let the bot win or lose for speed. The bot finds missing textures, alignment and rendering bugs, uncatalogued screens and real crashes, so it actually works and scales.
AI-written summary · about
chapters (6)
- 0:00Start
- 9:31Button detection - Result
- 12:40tection - Hash & Text similarity
- 14:20tection results
- 18:41console commands
- 23:13objects
related talks
46mHow King Uses AI in Testing
29mDragonCrawl: The AI Testing Framework That Adapts Instead of Breaks
35mApp Testing with AI Tools – Make it Easy
43mTesting Agentic AI Applications: Beyond Traditional QA
11mDragonCrawl: Revolutionizing Mobile Testing with GenAI
40mShould We Let AI Take Over Test Automation Completely?
19mQA Agents for Android: How AI Is Changing the Game in Testing
20mYour AI Wrote the Code. Now What?