Knowledge-infused Learning for Entity Prediction in Driving Scenes

Scene understanding is a key technical challenge within the autonomous driving domain. It requires a deep semantic understanding of the entities and relations found within complex physical and social environments that is both accurate and complete. In practice, this can be accomplished by representi...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Ruwan Wickramarachchi, Cory Henson , Amit Sheth
Formato:	article
Lenguaje:	EN
Publicado:	Frontiers Media S.A. 2021
Materias:	neuro-symbolic computing knowledge-infused learning knowledge graph embeddings autonomous driving scene understanding entity prediction Information technology T58.5-58.64
Acceso en línea:	https://doaj.org/article/730fa0d501bb4dfdae6aaf0508061c4d
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:730fa0d501bb4dfdae6aaf0508061c4d
record_format	dspace
spelling	oai:doaj.org-article:730fa0d501bb4dfdae6aaf0508061c4d2021-12-01T01:34:43ZKnowledge-infused Learning for Entity Prediction in Driving Scenes2624-909X10.3389/fdata.2021.759110https://doaj.org/article/730fa0d501bb4dfdae6aaf0508061c4d2021-11-01T00:00:00Zhttps://www.frontiersin.org/articles/10.3389/fdata.2021.759110/fullhttps://doaj.org/toc/2624-909XScene understanding is a key technical challenge within the autonomous driving domain. It requires a deep semantic understanding of the entities and relations found within complex physical and social environments that is both accurate and complete. In practice, this can be accomplished by representing entities in a scene and their relations as a knowledge graph (KG). This scene knowledge graph may then be utilized for the task of entity prediction, leading to improved scene understanding. In this paper, we will define and formalize this problem as Knowledge-based Entity Prediction (KEP). KEP aims to improve scene understanding by predicting potentially unrecognized entities by leveraging heterogeneous, high-level semantic knowledge of driving scenes. An innovative neuro-symbolic solution for KEP is presented, based on knowledge-infused learning, which 1) introduces a dataset agnostic ontology to describe driving scenes, 2) uses an expressive, holistic representation of scenes with knowledge graphs, and 3) proposes an effective, non-standard mapping of the KEP problem to the problem of link prediction (LP) using knowledge-graph embeddings (KGE). Using real, complex and high-quality data from urban driving scenes, we demonstrate its effectiveness by showing that the missing entities may be predicted with high precision (0.87 Hits@1) while significantly outperforming the non-semantic/rule-based baselines.Ruwan WickramarachchiCory Henson Amit Sheth Frontiers Media S.A.articleneuro-symbolic computingknowledge-infused learningknowledge graph embeddingsautonomous drivingscene understandingentity predictionInformation technologyT58.5-58.64ENFrontiers in Big Data, Vol 4 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	neuro-symbolic computing knowledge-infused learning knowledge graph embeddings autonomous driving scene understanding entity prediction Information technology T58.5-58.64
spellingShingle	neuro-symbolic computing knowledge-infused learning knowledge graph embeddings autonomous driving scene understanding entity prediction Information technology T58.5-58.64 Ruwan Wickramarachchi Cory Henson Amit Sheth Knowledge-infused Learning for Entity Prediction in Driving Scenes
description	Scene understanding is a key technical challenge within the autonomous driving domain. It requires a deep semantic understanding of the entities and relations found within complex physical and social environments that is both accurate and complete. In practice, this can be accomplished by representing entities in a scene and their relations as a knowledge graph (KG). This scene knowledge graph may then be utilized for the task of entity prediction, leading to improved scene understanding. In this paper, we will define and formalize this problem as Knowledge-based Entity Prediction (KEP). KEP aims to improve scene understanding by predicting potentially unrecognized entities by leveraging heterogeneous, high-level semantic knowledge of driving scenes. An innovative neuro-symbolic solution for KEP is presented, based on knowledge-infused learning, which 1) introduces a dataset agnostic ontology to describe driving scenes, 2) uses an expressive, holistic representation of scenes with knowledge graphs, and 3) proposes an effective, non-standard mapping of the KEP problem to the problem of link prediction (LP) using knowledge-graph embeddings (KGE). Using real, complex and high-quality data from urban driving scenes, we demonstrate its effectiveness by showing that the missing entities may be predicted with high precision (0.87 Hits@1) while significantly outperforming the non-semantic/rule-based baselines.
format	article
author	Ruwan Wickramarachchi Cory Henson Amit Sheth
author_facet	Ruwan Wickramarachchi Cory Henson Amit Sheth
author_sort	Ruwan Wickramarachchi
title	Knowledge-infused Learning for Entity Prediction in Driving Scenes
title_short	Knowledge-infused Learning for Entity Prediction in Driving Scenes
title_full	Knowledge-infused Learning for Entity Prediction in Driving Scenes
title_fullStr	Knowledge-infused Learning for Entity Prediction in Driving Scenes
title_full_unstemmed	Knowledge-infused Learning for Entity Prediction in Driving Scenes
title_sort	knowledge-infused learning for entity prediction in driving scenes
publisher	Frontiers Media S.A.
publishDate	2021
url	https://doaj.org/article/730fa0d501bb4dfdae6aaf0508061c4d
work_keys_str_mv	AT ruwanwickramarachchi knowledgeinfusedlearningforentitypredictionindrivingscenes AT coryhenson knowledgeinfusedlearningforentitypredictionindrivingscenes AT amitsheth knowledgeinfusedlearningforentitypredictionindrivingscenes
_version_	1718405966922252288

Knowledge-infused Learning for Entity Prediction in Driving Scenes

Ejemplares similares