I Put a Neural Network in a Thermometer — Then It Got Out of Hand
Alex Rosito shares his journey of integrating a neural network into a thermometer. Initially a simple project, it evolved into a tool called Hasaki that trains and exports neural networks as standalone C headers. This innovation allows for AI inference on microcontrollers not originally designed for such tasks.
- ▪Rosito began by building a basic thermometer using a thermistor and microcontroller.
- ▪He later replaced the polynomial correction function with a neural network, successfully running it on an ATtiny85 microcontroller.
- ▪Hasaki, the tool he developed, trains fully-connected feedforward networks and exports them without dependencies.
Opening excerpt (first ~120 words) tap to expand
try { if(localStorage) { let currentUser = localStorage.getItem('current_user'); if (currentUser) { currentUser = JSON.parse(currentUser); if (currentUser.id === 3903394) { document.getElementById('article-show-container').classList.add('current-user-is-article-author'); } } } } catch (e) { console.error(e); } Alex Rosito Posted on May 25 I Put a Neural Network in a Thermometer — Then It Got Out of Hand #cpp #embeddedsystems #machinelearning #iot How it all started About three years ago my wife asked me: "Honey, what do you think the temperature is today?" I told her to check her phone. She gave me a look and said, "No — in here. In the apartment." The next day I remembered we had a clock with a built-in thermometer somewhere. Found it, put batteries in it, problem solved.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at DEV.to (Top).
Discussion
0 commentsMore from DEV.to (Top)
