This system uses artificial intelligence to analyze video recordings of a person and objects to understand their physical movements and assess their function. It processes video frames through neural networks to identify and track the person and any relevant objects, detect their body positions, and recognize how they move. From this data, it calculates metrics to evaluate physical function, with the claims specifically detailing a method for tracking using feedforward and convolutional neural networks.
Why it matters: Since 2020, the performance and accessibility of pre-trained computer vision models and AI frameworks have significantly advanced. This makes implementing the described neural network-based tracking and assessment more efficient and accurate now.
AI gives you a few directions you could take this. Pick one, and we check whether your version is different enough to patent, then write the filing.
Reinvent this with AI