EdgeAI - AI in the Palm of Your Hand
summary
Claims that models run on-device for real time decisions and privacy, and demonstrates the actual toolchain: TensorFlow Lite, ML Kit/Firebase ML, Coral Edge TPU, Jetson Nano and Arduino/SparkFun Edge. Gives surprising numbers: 250 billion microcontrollers in the world, 28B shipped in 2018 and 38B by 2023. Shows things built: a Flutter face‑detection mustache face demo, a tiny‑ML yes/no model shrunk to 20 kilobits running on embedded boards, and an Arduino gesture demo with measured accuracies.
AI-written summary · about
chapters (19)
- 0:00Intro
- 1:06Etiquette
- 5:00Sponsor
- 6:30Introduction
- 7:30Know your audience
- 8:30Where do your models run
- 9:45What is EdgeAI
- 11:40Devices with EdgeAI
- 13:15Why EdgeAI
- 16:25Mobile Use Cases
- 18:40Firebase ML Kit
- 21:40Face Detection App
- 23:10Microprocessors and ML
- 25:25Microprocessor Overview
- 30:21Raspberry Pi
- 36:13Considerations
- 38:02Demo
- 45:04QA
- 47:19Questions
related talks
21mUnlocking the potential of low-power embedded devices with Edge AI software
10mThe Rise and Impact of Edge AI
22mEdge AI Use Cases in IoT — Unlocking Intelligence at the Network Edge
01h 01mML and the IoT: Living on the Edge
23m2025: On‑Device AI Goes Mainstream on Android
50mML and AI at the edge with Google Coral, EdgeTPU and Cloud Platform
48mThe Power of Edge AI: Revolutionizing Future of Real-Time, Autonomous Processing
31mKeynote: The next big thing in machine learning is tiny