Motivated by neighbor’s cat that poops on his lawn, an NVIDIA engineer built a smart system in less than fifteen hours that automatically turns on the sprinklers when a cat is on his lawn.
Robert Bond, a system software engineer, first trained a neural network to recognize cat images using a TITAN GPU and the cuDNN-accelerated Caffe deep learning framework.
He then set up an NVIDIA Jetson TX1 board with a small camera to feed video frames through the neural network. When it recognizes a cat, the Jetson signals a cloud server which then responds by turning on the sprinklers for two minutes.
“It wasn’t actually that much work,” Bond says. “The new Jetson TX1 is really good at running these neural nets.”
Bond details about choosing and training the neural network as well as setting up the hardware so you can get started on similar projects yourself.
The next challenge for Bond is to not just detect cats, but detect their location, and automatically send a remote-controlled car to shoo the critters away.
Read more >>
AI-Generated Summary
- NVIDIA system software engineer Robert Bond built an automated sprinkler system that detects cats using a neural network trained on a neural network with a TITAN GPU and cuDNN-accelerated Caffe deep learning framework.
- An NVIDIA Jetson TX1 board processes video frames from a camera through the trained network and signals a cloud server to activate sprinklers for two minutes when a cat is recognized.
- Bond plans to extend the system to detect cat locations and deploy a remote-controlled vehicle to deter them.
Next Steps
- Read the AI Cat Chaser project details to learn how to train the neural network and set up the hardware.
Powered by NVIDIA Nemotron. AI-generated content may summarize information incompletely. Verify important information. Learn more
About the Authors
