Keynote: The next big thing in machine learning is tiny
summary
Claims that ML can run on tiny battery-powered devices, citing ~3 billion capable devices and a prediction of 8.4 billion voice assistants; demonstrates 13 KB wake-word model and ~20 KB object detector. Shows a working aquaponics monitor built and deployed to a microcontroller in about two hours that reliably detects running water, using tools like Edge Impulse and TensorFlow Lite (with ONNX for conversion). Notes dev boards cost ~$50–60 and simple sensor+MCU builds can be $10–15 and run years on a battery.
AI-written summary · about
chapters (6)
- 0:00Start
- 7:24Data centric machine learning
- 8:10The extremes of machine learning
- 18:04Aquaponics system
- 25:02Edge and cloud working together
- 25:51Augmented Reality & Virtual Reality
related talks
01h 53mTinyML: Machine Learning’s Future is Tiny and Bright
27mWhat’s TinyML good for
25mMachine Learning on Microcontrollers; Tiny, Useful and Fun
43mKeynote: Tiny Data Collectors, Vastly (...) & the Land of Tiny Challenges
55mEdgeAI - AI in the Palm of Your Hand
01h 01mML and the IoT: Living on the Edge
54mSmarter Everything with ML & the IoT
22mAI On The Farthest Edge