Image processing method and apparatus
AssigneesThe Board of Trustees of the University of Illinois (Urbana, IL), Samsung Electronics Co., Ltd.
InventorsAlexander Schwing, Colin Graber, Raymond Yeh, Jihye Kim, Jaejoon Han
Life of this patent
Estimating the term…
Abstract
An image processing method includes defining relations between entities of a target of which a motion is to be predicted from an image of a first time point based on a feature vector of the entities, estimating a dynamic interaction between the entities at the first time point based on the defined relations between the entities, predicting a motion of the entities changing at a second time point based on the estimated dynamic interaction, and outputting a result to which the motion predicted at the second time point is applied.
Claims19
An image processing method comprising:
- defining relations between entities of a target of which a motion is to be predicted from an image of a first time point based on a feature vector of the entities;
- estimating a dynamic interaction between the entities at the first time point based on the defined relations between the entities, including sampling a result in which prior information and posterior information are combined and generating a latent variable corresponding to the dynamic interaction between the entities at the first time point based on a result of the sampling;
- predicting a motion of the entities changing at a second time point based on the estimated dynamic interaction and the latent variable;
- and
- outputting a result to which the motion predicted at the second time point is applied.
An image processing apparatus comprising:
- a communication interface configured to receive an image of a first time point comprising entities of a target of which a motion is to be predicted;
- a processor configured to extract a feature vector of the entities from the image of the first timepoint, to estimate a dynamic interaction between the entities at the first time point based on relations between the entities defined based on the feature vector, including sampling a result in which prior information and posterior information are combined and generating a latent variable corresponding to the dynamic interaction between the entities at the first time point based on a result of the sampling, and to predict a motion of the entities changing at a second time point based on the estimated dynamic interaction and the latent variable;
- and
- an output device configured to output a result to which the motion predicted at the second time point is applied.
Dependencies are inferred from the claim text.