" Definitely need this app in my life! Congrats on the launch, really excited to try it out and see if it has any effect on my back pain. ”
“Thanks for making this available. I have horrible posture and am excited to try this out!”
Latest technologies and breakthroughs in machine learning and computer vision
All you need is a Mac (Big Sur) with a front camera.
Your posture data never leaves your device, period.
Accurate detection based on your unique posture.
On-device Learning, Machine Learning, Computer Vision.
Up and running in less than a minute.
You don't even need the internet because the model is trained on-device.
Good posture. Sit straight up, eyes one arm away from your computer.
Bad posture. Slouching causes neck pain, back pain, dry eyes and even headaches.
Many people have started using PostureNet to improve their posture.
Hours of research
It requires a bit of explanation, in high level, PostureNet is a deep neural network based on the latest computer vision and AI technologies.
Yes, right now, the app only runs on Mac (Big Sur).
Posture Net takes advantage of many advanced ML features that the Mac platform provides to us such as Core ML and computer vision.
If you would like to see Posture Net on other platforms, please feel free to let us know.
To use Posture Net you need:
1. A Mac that runs the latest macOS 11(Big Sur).
2. The Mac is equipped with a front camera.
First of all, none of your posture data leaves your machine.
We do not collect any of your posture data, nor did we share it with anyone.
But the app does need to understand your unique posture to be able to remind you when your posture is bad.
You are basically teaching your Mac how to differentiate good postures from bad postures.
Posture Net works best when your computer camera is in a 'stable' environment.
If you switched to another environment, e.g., moved from office to home or sitting to standing, the app may still think you are in the previous setting and not perform as well as before.
When you switch to a new environment or notice the detection is inaccurate, most often, it's because your postured data had become different than when you first configured it.
That's when it is a good time to reset it.