Inseer Inc.

University of Iowa Research Foundation

John Rachid

Jean E Robillard

Mitchell Messmore

Alec Diaz-Arias

Dmitriy Shin

Stephen Baek

Dmitry Shin

Aiec Diaz-Arias

An apparatus for 3D human pose estimation using dynamic multi-headed convolutional attention mechanism is presented. The apparatus contains two dynamic multi-headed convolutional attention mechanism with spatial attention and another with temporal attention that leverages the spatial attention mechanism to extract frame-wise inter-joint dependencies by analyzing sections of limbs that are related. The temporal attention mechanism extracts global inter-frame relationships by analyzing correlations between the temporal profile of joints. The temporal profile mechanism leads to a more diverse temporal attention map while achieving substantial parameter reduction.

Methods and apparatus for human pose estimation from images using dynamic multi-headed convolutional attention

A method can include receiving (1) images of at least one subject and (2) at least one total mass value for the at least one subject. The method can further include executing a first machine learning model to identify joints of the at least one subject. The method can further include executing a second machine learning model to determine limbs of the at least one subject based on the joints and the images. The method can further include generating three-dimensional (3D) representations of a skeleton based on the joints and the limbs. The method can further include determining a torque value for each limb, based on at least one of a mass value and a linear acceleration value, or a torque inertia and an angular acceleration value. The method can further include generating a risk assessment report based on at least one torque value being above a predetermined threshold.

Methods and apparatus for machine learning to analyze musculo-skeletal rehabilitation from images

An apparatus for calculating torque and force about body joints to predict muscle fatigue includes a processor configured to receive image frames depicting a subject. The processor is configured to execute at least one machine learning model using the image frames as an input, to generate a 2D representation of the subject, a subject mass value for the subject based on the 2D representation, and a 3D representation of the subject based on the 2D representation, where the 3D representation includes a temporal joints profile. The processor is further configured to compute each torque value for each joint of the subject from the 3D representation, based on the subject mass value. The processor is further configured to generate a muscle fatigue prediction for each joint of the subject, based on a set of torque values and a torque threshold.

Methods and apparatus for calculating torque and force about body joints using machine learning to predict muscle fatigue

Growing community of inventors

Iowa City, IA, United States of America

John Rachid