Synergistic Object Tracking and Pattern Recognition for Event Representation

Number: 20230252651
Published: 2023-08-10
Filed: 2023-04-04
Assignee: Disney Enterprises, Inc.
Inventors: Comito; Keith et al.
CPC: G06T7/292; G06T7/251
Verdict: Set aside generic tracking/analytics, no creative hook
Source: Google Patents · FreePatentsOnline

Abstract

A system for performing synergistic object tracking and pattern recognition for event representation includes a computing platform having processing hardware and a system memory storing a software code. The processing hardware is configured to execute the software code to receive event data corresponding to one or more propertie(s) of an object, to generate, using the event data, a location data estimating a location of each of multiple predetermined landmarks of the object, and to predict, using one or both of the event data and the location data, a pattern corresponding to the propertie(s) of the object. The processing hardware is further configured to execute the software code to update, using the predicted pattern, the location data, and to merge the updated location data and the predicted pattern to provide merged data.

Background

BACKGROUND

Computer vision systems can perform poorly when a body is moving dynamically or in an atypical manner, as well as when multiple bodies are in close proximity with one another, or are in physical contact. For example, during grappling between competitors in a wrestling, judo, or mixed martial arts competition, a system may be unable to generate accurate three-dimensional (3D) skeletal data when the bodies of the competitors occlude one another. This may significantly limit the ability to identify the full action sequence during a physically interactive event, such as a combat sports event or a dance competition, for example.

Claims

21. A system comprising: a computing platform including a processing hardware and a system memory storing a software code; the processing hardware configured to execute the software code to: receive a video including physical interactions between a first human and a first object, wherein the first object occludes at least a portion of the first human in the video; generate, using the video, a location data representing a location of each of a plurality of skeletal landmarks of the first human, wherein the location data include an estimated location of each of at least one or more of the plurality of skeletal landmarks of the first human occluded by the first object; predict, using at least one of the video or the location data, a pattern corresponding to one or more of movement, location or posture of the first human; update, using the predicted pattern, the location data to provide an updated location data; and merge the updated location data and the predicted pattern to provide merged data. || 31. A computer-implemented method comprising: receiving a video including physical interactions between a first human and a first object, wherein the first object occludes at least a portion of the first human in the video; generating, using the video, a location data representing a location of each of a plurality of skeletal landmarks of the first human, wherein the location data include an estimated location of each of at least one or more of the plurality of skeletal landmarks of the first human occluded by the first object; predicting, using at least one of the video or the location data, a pattern corresponding to one or more of movement, location or posture of the first human; updating, using the predicted pattern, the location data to provide an updated location data; and merging the updated location data and the predicted pattern to provide merged data.