Representation and retrieval of images by means of spatial relations between objects

Danilo Nunes; Leonardo Anjoletto Ferreira; Paulo Eduardo Santos; Adam Pease

Representation and retrieval of images by means of spatial relations between objects

Danilo Nunes, Leonardo Anjoletto Ferreira, Paulo Eduardo Santos, Adam Pease

Research output: Contribution to journal › Conference article › peer-review

35 Downloads (Pure)

Abstract

The present work addresses the challenge of integrating low-level information with high-level knowledge (known as semantic gap) that exists in content-based image retrieval by introducing an approach to describe images by means of spatial relations. The proposed approach is called Image Retrieval using Region Analysis (IRRA) and relies on decomposing images into pairs of objects. This method generates a representation composed of n triples, each one containing: a noun, a preposition and, another noun. This representation paves the way to enable image retrieval based on spatial relations. Results for an indoor/outdoor classifier shows that neural networks alone are capable of achieving 88% in precision and recall, but when combined with ontology this result increases in 10 percentage points, reaching 98% of precision and recall.

Original language	English
Journal	CEUR Workshop Proceedings
Volume	2350
Publication status	Published - 1 Jan 2019
Externally published	Yes
Event	2019 AAAI Spring Symposium on Combining Machine Learning with Knowledge Engineering, AAAI-MAKE 2019 - Palo Alto, United States Duration: 25 Mar 2019 → 27 Mar 2019

Keywords

Image Retrieval Region Analysis (IRRA)
spatial relations

Access to Document

Nunes_Representation_P2019Final published version, 891 KBLicence: In Copyright

http://ceur-ws.org/Vol-2350/Licence: In Copyright

Cite this

@article{856b7637e21b40e7922499da2a8264f8,

title = "Representation and retrieval of images by means of spatial relations between objects",

abstract = "The present work addresses the challenge of integrating low-level information with high-level knowledge (known as semantic gap) that exists in content-based image retrieval by introducing an approach to describe images by means of spatial relations. The proposed approach is called Image Retrieval using Region Analysis (IRRA) and relies on decomposing images into pairs of objects. This method generates a representation composed of n triples, each one containing: a noun, a preposition and, another noun. This representation paves the way to enable image retrieval based on spatial relations. Results for an indoor/outdoor classifier shows that neural networks alone are capable of achieving 88% in precision and recall, but when combined with ontology this result increases in 10 percentage points, reaching 98% of precision and recall.",

keywords = "Image Retrieval Region Analysis (IRRA), spatial relations",

author = "Danilo Nunes and Ferreira, {Leonardo Anjoletto} and Santos, {Paulo Eduardo} and Adam Pease",

year = "2019",

month = jan,

day = "1",

language = "English",

volume = "2350",

journal = "CEUR Workshop Proceedings",

issn = "1613-0073",

publisher = "CEUR-WS",

note = "2019 AAAI Spring Symposium on Combining Machine Learning with Knowledge Engineering, AAAI-MAKE 2019 ; Conference date: 25-03-2019 Through 27-03-2019",

}

TY - JOUR

T1 - Representation and retrieval of images by means of spatial relations between objects

AU - Nunes, Danilo

AU - Ferreira, Leonardo Anjoletto

AU - Santos, Paulo Eduardo

AU - Pease, Adam

PY - 2019/1/1

Y1 - 2019/1/1

N2 - The present work addresses the challenge of integrating low-level information with high-level knowledge (known as semantic gap) that exists in content-based image retrieval by introducing an approach to describe images by means of spatial relations. The proposed approach is called Image Retrieval using Region Analysis (IRRA) and relies on decomposing images into pairs of objects. This method generates a representation composed of n triples, each one containing: a noun, a preposition and, another noun. This representation paves the way to enable image retrieval based on spatial relations. Results for an indoor/outdoor classifier shows that neural networks alone are capable of achieving 88% in precision and recall, but when combined with ontology this result increases in 10 percentage points, reaching 98% of precision and recall.

AB - The present work addresses the challenge of integrating low-level information with high-level knowledge (known as semantic gap) that exists in content-based image retrieval by introducing an approach to describe images by means of spatial relations. The proposed approach is called Image Retrieval using Region Analysis (IRRA) and relies on decomposing images into pairs of objects. This method generates a representation composed of n triples, each one containing: a noun, a preposition and, another noun. This representation paves the way to enable image retrieval based on spatial relations. Results for an indoor/outdoor classifier shows that neural networks alone are capable of achieving 88% in precision and recall, but when combined with ontology this result increases in 10 percentage points, reaching 98% of precision and recall.

KW - Image Retrieval Region Analysis (IRRA)

KW - spatial relations

UR - http://www.scopus.com/inward/record.url?scp=85064993481&partnerID=8YFLogxK

M3 - Conference article

AN - SCOPUS:85064993481

SN - 1613-0073

VL - 2350

JO - CEUR Workshop Proceedings

JF - CEUR Workshop Proceedings

T2 - 2019 AAAI Spring Symposium on Combining Machine Learning with Knowledge Engineering, AAAI-MAKE 2019

Y2 - 25 March 2019 through 27 March 2019

ER -

Representation and retrieval of images by means of spatial relations between objects

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this