Our deep learning models can work with any wearable and seamlessly count and classify any motion pattern in real-time. We also consider motion repetition timing and consistency as a key metric in analyzing the quality of the motion and provide that data in real-time too
We can use a phone's camera to count and time any motion pattern repetitions in real-time. The models run on the device itself and provide real-time feedback to the user
We have state-of-the-art models that work with any camera or any wearable like a smart watch
In depth real-time analysis including repetition counting, repetition timing and motion classification
Our AI models can provide real-time analysis for any motion pattern. No user training is needed
All the AI models run on the device which could be a phone or a smartwatch e.g. our wearable models run on Apple Watch
We have built state-of-the-art deep learning models which seamlessly capture motion analytics using either a wearable like smartwatch or a phone’s camera. The analytics can provide real-time details like repetition counting, repetition timing and consistency and even recognize several different types of motion
Our models work for most exercises and no pre-training or exercise library is required. The models work for any camera or any wearable out of the box
Yes, there is no interdependence between the two technologies and they can be deployed independently.
No cloud server is needed. All the models – including for wearable devices and camera run on the devices (phone/watch) in real-time, so you can deploy them at the edge
The models have been trained on 100k+ labeled repetition data collected from real users in collaboration with over 50 personal trainers over many months. The data has been cleaned and verified and our models have over 95%+ accuracy on the test data set.
Accuracy of our models is +/- 1 repetition for most exercises. For wearable technology, there are some cases where the accuracy can be poorer e.g. when the battery of devices is low or when there is a poor connection between wearable and phone
© 2021. All rights reserved.