TinyML - efficient deep learning on edge & embedded devices
Created: 29 Dec 2022, 10:45 AM | Modified: =dateformat(this.file.mtime,"dd MMM yyyy, hh:mm a")
Tags: knowledge, TinyML
Blogpost, Medium post, allaboutcircuits article, about TinyML
The TinyMLproject aims to improve the efficiency of deep learning AI systems by requiring less computation, fewer engineers, and less data, to facilitate the giant market of edge AI and AIoT.
From <https://tinyml.mit.edu/>
- Led by Song Han at MIT HAN Lab
- MCUNet (slides overview, MCUNetv2 overview), Once-for-all NAS method
Courses
- http://tinyml.seas.harvard.edu/
- Song Han: MIT course: TinyML and Efficient Deep Learning
tinyML Talks Singapore Meet and Greet
https://github.com/VincentK16/tinymlresources
Papers: TinyML for Concept Drift TinyOL - TinyML with Online-Learning on Microcontrollers MIT 2022 - TinyML EfficientML Course (Prof Song Han)