DOE Patents title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Universal image representation based on a multimodal graph

Abstract

A system for classifying a target image with segments having attributes is provided. The system generates a graph for the target image that includes vertices representing segments of the image and edges representing relationships between the connected vertices. For each vertex, the system generates a subgraph that includes the vertex as a home vertex and neighboring vertices representing segments of the target image within a neighborhood of the segment represented by the home vertex. The system applies an autoencoder to each subgraph to generate latent variables to represent the subgraph. The system applies a machine learning algorithm to a feature vector comprising a universal image representation of the target image that is derived from the generated latent variables of the subgraphs to generate a classification for the target image.

Inventors:
; ;
Issue Date:
Research Org.:
Lawrence Livermore National Laboratory (LLNL), Livermore, CA (United States)
Sponsoring Org.:
USDOE
OSTI Identifier:
2222186
Patent Number(s):
11735311
Application Number:
17/470,331
Assignee:
Lawrence Livermore National Security, LLC (Livermore, CA)
DOE Contract Number:  
AC52-07NA27344
Resource Type:
Patent
Resource Relation:
Patent File Date: 09/09/2021
Country of Publication:
United States
Language:
English

Citation Formats

Bremer, Peer-Timo, Anirudh, Rushil, and Thiagarajan, Jayaraman Jayaraman. Universal image representation based on a multimodal graph. United States: N. p., 2023. Web.
Bremer, Peer-Timo, Anirudh, Rushil, & Thiagarajan, Jayaraman Jayaraman. Universal image representation based on a multimodal graph. United States.
Bremer, Peer-Timo, Anirudh, Rushil, and Thiagarajan, Jayaraman Jayaraman. Tue . "Universal image representation based on a multimodal graph". United States. https://www.osti.gov/servlets/purl/2222186.
@article{osti_2222186,
title = {Universal image representation based on a multimodal graph},
author = {Bremer, Peer-Timo and Anirudh, Rushil and Thiagarajan, Jayaraman Jayaraman},
abstractNote = {A system for classifying a target image with segments having attributes is provided. The system generates a graph for the target image that includes vertices representing segments of the image and edges representing relationships between the connected vertices. For each vertex, the system generates a subgraph that includes the vertex as a home vertex and neighboring vertices representing segments of the target image within a neighborhood of the segment represented by the home vertex. The system applies an autoencoder to each subgraph to generate latent variables to represent the subgraph. The system applies a machine learning algorithm to a feature vector comprising a universal image representation of the target image that is derived from the generated latent variables of the subgraphs to generate a classification for the target image.},
doi = {},
journal = {},
number = ,
volume = ,
place = {United States},
year = {2023},
month = {8}
}

Works referenced in this record:

Object pose estimation
patent-application, November 2020


Learning Property Graph Representations Edge-By- Edge
patent-application, May 2020


Categorical feature encoding for property graphs by vertex proximity
patent-application, August 2020


Apparatus and method to process and cluster data
patent-application, March 2019