Launching TensorFlow Lite for Microcontrollers

petewarden.com

Pete Warden is on the TensorFlow Lite team at Google and has written quite a lot of great stuff on the intersection of IoT and deep learning. This post is a transcription of a conference talk he gave in which he demos a speech recognition model using TF lite that takes up a whopping 20K of RAM and detects the word "yes".

This space is still very early, but it's tremendously exciting to see what's happening in it. Most of the industry is still thinking about plugged in, cloud-connected applications of data science, but that's going to evolve...

Read more...
Linkedin

Want to receive more content like this in your inbox?