Graph Signal Processing for Scene Representation and Analysis /Author=Tian, Dong; Mansour, Hassan; Cohen, Robert A.; Vetro, Anthony /CreationDate=May 25, 2016 /Subject=Digital Video

Graph signal processing (GSP) is an emerging field that provides a new family of tools for analyzing signals that could be modeled on vertices connected by edges. In this paper, we describe two examples of how GSP is being applied for scene representation and analysis, where the scene is either captured as video sequences or point clouds. In the first example, we show that novel graph constructions can be used to robustly segment moving foreground objects from the background of video sequences with ego-motions. In the second example, we employ a graph-based transform to efficiently code attributes associated with the point clouds. We demonstrate with the two examples the potential benefits of using GSP tools for scene representation and analysis. Graph Signal Processing Workshop (GSP) This work may not be copied or reproduced in whole or in part for any commercial purpose. Permission to copy in whole or in part without payment of fee is granted for nonprofit educational and research purposes provided that all such whole or partial copies include the following: a notice that such copying is by permission of Mitsubishi Electric Research Laboratories, Inc.; an acknowledgment of the authors and individual contributions to the work; and all applicable portions of the copyright notice. Copying, reproduction, or republishing for any other purpose shall require a license with payment of fee to Mitsubishi Electric Research Laboratories, Inc. All rights reserved. Copyright c © Mitsubishi Electric Research Laboratories, Inc., 2016 201 Broadway, Cambridge, Massachusetts 02139 Graph Signal Processing for Scene Representation and Analysis Dong Tian, Hassan Mansour, Robert Cohen, Anthony Vetro Mitsubishi Electronic Research Labs (MERL) 201 Broadway, Cambridge, MA 02139, USA {tian; mansour; cohen; avetro}@merl.com Abstract—Graph signal processing (GSP) is an emerging field that provides a new family of tools for analyzing signals that could be modeled on vertices connected by edges. In this paper, we describe two examples of how GSP is being applied for scene representation and analysis, where the scene is either captured as video sequences or point clouds. In the first example, we show that novel graph constructions can be used to robustly segment moving foreground objects from the background of video sequences with ego-motions. In the second example, we employ a graph-based transform to efficiently code attributes associated with the point clouds. We demonstrate with the two examples the potential benefits of using GSP tools for scene representation and analysis.Graph signal processing (GSP) is an emerging field that provides a new family of tools for analyzing signals that could be modeled on vertices connected by edges. In this paper, we describe two examples of how GSP is being applied for scene representation and analysis, where the scene is either captured as video sequences or point clouds. In the first example, we show that novel graph constructions can be used to robustly segment moving foreground objects from the background of video sequences with ego-motions. In the second example, we employ a graph-based transform to efficiently code attributes associated with the point clouds. We demonstrate with the two examples the potential benefits of using GSP tools for scene representation and analysis.