17987460. METRIC LIFTING OF 3D HUMAN POSE USING SOUND simplified abstract (SAMSUNG ELECTRONICS CO., LTD.)

From WikiPatents
Jump to navigation Jump to search

METRIC LIFTING OF 3D HUMAN POSE USING SOUND

Organization Name

SAMSUNG ELECTRONICS CO., LTD.

Inventor(s)

Zhijian Yang of Urbana IL (US)

Xiaoran Fan of Irvine CA (US)

Ibrahim Volkan Isler of Saint Paul MN (US)

Hyun Soo Park of New York NY (US)

METRIC LIFTING OF 3D HUMAN POSE USING SOUND - A simplified explanation of the abstract

This abstract first appeared for US patent application 17987460 titled 'METRIC LIFTING OF 3D HUMAN POSE USING SOUND

Simplified Explanation

Abstract: A patent application describes a method for estimating the pose of a person using an image and audio impulse responses. The image represents a 2D scene with the person, while the audio impulse responses are obtained in different environmental conditions. The pose is reconstructed by combining the image and audio impulse responses, resulting in a metric scale human pose estimation.

  • The patent application proposes a method for estimating the pose of a person.
  • The pose estimation is based on an image and audio impulse responses.
  • The image represents a 2D scene containing the person.
  • Audio impulse responses are obtained in different environmental conditions.
  • The pose is reconstructed by combining the image and audio impulse responses.
  • The resulting pose estimation is in metric scale, providing accurate measurements of the person's pose.

Potential Applications:

  • Human-computer interaction systems that require accurate pose estimation.
  • Virtual reality and augmented reality applications for realistic avatar representation.
  • Sports analysis and training systems for tracking and analyzing athletes' movements.
  • Surveillance systems for identifying and tracking individuals based on their pose.
  • Medical applications for monitoring patients' movements and posture.

Problems Solved:

  • Accurate and reliable estimation of a person's pose using a combination of image and audio data.
  • Overcoming limitations of traditional pose estimation methods that rely solely on visual information.
  • Providing metric scale pose estimation for precise measurements and analysis.

Benefits:

  • Improved accuracy and reliability in estimating a person's pose.
  • Enhanced realism in virtual reality and augmented reality applications.
  • More comprehensive analysis and tracking of athletes' movements in sports.
  • Improved identification and tracking capabilities in surveillance systems.
  • Better monitoring and analysis of patients' movements and posture in medical applications.


Original Abstract Submitted

A pose of a person is estimated using an image and audio impulse responses. The image represents a 2D scene including the person. The audio impulse responses are obtained with the present absent and present in an environment. The pose is reconstructed based on the image and the one or more audio impulse responses. The pose is a metric scale human pose.