CubiCasa5K: A Dataset and an Improved Multi-task Model for Floorplan Image Analysis

Ahti Kalervo; Juha Ylioinas; Markus Häikiö; Antti Karhu; Juho Kannala

doi:10.1007/978-3-030-20205-7_3

CubiCasa5K: A Dataset and an Improved Multi-task Model for Floorplan Image Analysis

Ahti Kalervo^*
, Juha Ylioinas
, Markus Häikiö
, Antti Karhu
, Juho Kannala

^*Corresponding author for this work

CubiCasa Oy

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific › peer-review

124 Citations (Scopus)

Abstract

Better understanding and modelling of building interiors and the emergence of more impressive AR/VR technology has brought up the need for automatic parsing of floorplan images. However, there is a clear lack of representative datasets to investigate the problem further. To address this shortcoming, this paper presents a novel image dataset called CubiCasa5K, a large-scale floorplan image dataset containing 5000 samples annotated into over 80 floorplan object categories. The dataset annotations are performed in a dense and versatile manner by using polygons for separating the different objects. Diverging from the classical approaches based on strong heuristics and low-level pixel operations, we present a method relying on an improved multi-task convolutional neural network. By releasing the novel dataset and our implementations, this study significantly boosts the research on automatic floorplan image analysis as it provides a richer set of tools for investigating the problem in a more comprehensive manner. Data and code at: https://github.com/CubiCasa/CubiCasa5k.

Original language	English
Title of host publication	Image Analysis - 21st Scandinavian Conference, SCIA 2019, Proceedings
Editors	Michael Felsberg, Per-Erik Forssén, Jonas Unger, Ida-Maria Sintorn
Publisher	Springer
Pages	28-40
Number of pages	13
ISBN (Print)	9783030202040
DOIs	https://doi.org/10.1007/978-3-030-20205-7_3
Publication status	Published - 1 Jan 2019
MoE publication type	A4 Conference publication
Event	Scandinavian Conference on Image Analysis - Norrköping, Sweden Duration: 11 Jun 2019 → 13 Jun 2019 Conference number: 21

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	11482 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	Scandinavian Conference on Image Analysis
Abbreviated title	SCIA
Country/Territory	Sweden
City	Norrköping
Period	11/06/2019 → 13/06/2019

Keywords

Convolutional neural networks
Dataset
Floorplan images
Multi-task learning

Access to Document

10.1007/978-3-030-20205-7_3

https://arxiv.org/abs/1904.01920

Cite this

Kalervo, A., Ylioinas, J., Häikiö, M., Karhu, A., & Kannala, J. (2019). CubiCasa5K: A Dataset and an Improved Multi-task Model for Floorplan Image Analysis. In M. Felsberg, P.-E. Forssén, J. Unger, & I.-M. Sintorn (Eds.), Image Analysis - 21st Scandinavian Conference, SCIA 2019, Proceedings (pp. 28-40). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11482 LNCS). Springer. https://doi.org/10.1007/978-3-030-20205-7_3

Kalervo, Ahti ; Ylioinas, Juha ; Häikiö, Markus et al. / CubiCasa5K : A Dataset and an Improved Multi-task Model for Floorplan Image Analysis. Image Analysis - 21st Scandinavian Conference, SCIA 2019, Proceedings. editor / Michael Felsberg ; Per-Erik Forssén ; Jonas Unger ; Ida-Maria Sintorn. Springer, 2019. pp. 28-40 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{7f2faf1c1da343e28e375b9e8deba2f4,

title = "CubiCasa5K: A Dataset and an Improved Multi-task Model for Floorplan Image Analysis",

abstract = "Better understanding and modelling of building interiors and the emergence of more impressive AR/VR technology has brought up the need for automatic parsing of floorplan images. However, there is a clear lack of representative datasets to investigate the problem further. To address this shortcoming, this paper presents a novel image dataset called CubiCasa5K, a large-scale floorplan image dataset containing 5000 samples annotated into over 80 floorplan object categories. The dataset annotations are performed in a dense and versatile manner by using polygons for separating the different objects. Diverging from the classical approaches based on strong heuristics and low-level pixel operations, we present a method relying on an improved multi-task convolutional neural network. By releasing the novel dataset and our implementations, this study significantly boosts the research on automatic floorplan image analysis as it provides a richer set of tools for investigating the problem in a more comprehensive manner. Data and code at: https://github.com/CubiCasa/CubiCasa5k.",

keywords = "Convolutional neural networks, Dataset, Floorplan images, Multi-task learning",

author = "Ahti Kalervo and Juha Ylioinas and Markus H{\"a}iki{\"o} and Antti Karhu and Juho Kannala",

year = "2019",

month = jan,

day = "1",

doi = "10.1007/978-3-030-20205-7\_3",

language = "English",

isbn = "9783030202040",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer",

pages = "28--40",

editor = "Michael Felsberg and Per-Erik Forss{\'e}n and Jonas Unger and Ida-Maria Sintorn",

booktitle = "Image Analysis - 21st Scandinavian Conference, SCIA 2019, Proceedings",

address = "Germany",

note = "Scandinavian Conference on Image Analysis, SCIA ; Conference date: 11-06-2019 Through 13-06-2019",

}

Kalervo, A, Ylioinas, J, Häikiö, M, Karhu, A & Kannala, J 2019, CubiCasa5K: A Dataset and an Improved Multi-task Model for Floorplan Image Analysis. in M Felsberg, P-E Forssén, J Unger & I-M Sintorn (eds), Image Analysis - 21st Scandinavian Conference, SCIA 2019, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11482 LNCS, Springer, pp. 28-40, Scandinavian Conference on Image Analysis, Norrköping, Sweden, 11/06/2019. https://doi.org/10.1007/978-3-030-20205-7_3

CubiCasa5K: A Dataset and an Improved Multi-task Model for Floorplan Image Analysis. / Kalervo, Ahti; Ylioinas, Juha; Häikiö, Markus et al.
Image Analysis - 21st Scandinavian Conference, SCIA 2019, Proceedings. ed. / Michael Felsberg; Per-Erik Forssén; Jonas Unger; Ida-Maria Sintorn. Springer, 2019. p. 28-40 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11482 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific › peer-review

TY - GEN

T1 - CubiCasa5K

T2 - Scandinavian Conference on Image Analysis

AU - Kalervo, Ahti

AU - Ylioinas, Juha

AU - Häikiö, Markus

AU - Karhu, Antti

AU - Kannala, Juho

N1 - Conference code: 21

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Better understanding and modelling of building interiors and the emergence of more impressive AR/VR technology has brought up the need for automatic parsing of floorplan images. However, there is a clear lack of representative datasets to investigate the problem further. To address this shortcoming, this paper presents a novel image dataset called CubiCasa5K, a large-scale floorplan image dataset containing 5000 samples annotated into over 80 floorplan object categories. The dataset annotations are performed in a dense and versatile manner by using polygons for separating the different objects. Diverging from the classical approaches based on strong heuristics and low-level pixel operations, we present a method relying on an improved multi-task convolutional neural network. By releasing the novel dataset and our implementations, this study significantly boosts the research on automatic floorplan image analysis as it provides a richer set of tools for investigating the problem in a more comprehensive manner. Data and code at: https://github.com/CubiCasa/CubiCasa5k.

AB - Better understanding and modelling of building interiors and the emergence of more impressive AR/VR technology has brought up the need for automatic parsing of floorplan images. However, there is a clear lack of representative datasets to investigate the problem further. To address this shortcoming, this paper presents a novel image dataset called CubiCasa5K, a large-scale floorplan image dataset containing 5000 samples annotated into over 80 floorplan object categories. The dataset annotations are performed in a dense and versatile manner by using polygons for separating the different objects. Diverging from the classical approaches based on strong heuristics and low-level pixel operations, we present a method relying on an improved multi-task convolutional neural network. By releasing the novel dataset and our implementations, this study significantly boosts the research on automatic floorplan image analysis as it provides a richer set of tools for investigating the problem in a more comprehensive manner. Data and code at: https://github.com/CubiCasa/CubiCasa5k.

KW - Convolutional neural networks

KW - Dataset

KW - Floorplan images

KW - Multi-task learning

UR - https://www.scopus.com/pages/publications/85066882608

U2 - 10.1007/978-3-030-20205-7_3

DO - 10.1007/978-3-030-20205-7_3

M3 - Conference article in proceedings

AN - SCOPUS:85066882608

SN - 9783030202040

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 28

EP - 40

BT - Image Analysis - 21st Scandinavian Conference, SCIA 2019, Proceedings

A2 - Felsberg, Michael

A2 - Forssén, Per-Erik

A2 - Unger, Jonas

A2 - Sintorn, Ida-Maria

PB - Springer

Y2 - 11 June 2019 through 13 June 2019

ER -

Kalervo A, Ylioinas J, Häikiö M, Karhu A, Kannala J. CubiCasa5K: A Dataset and an Improved Multi-task Model for Floorplan Image Analysis. In Felsberg M, Forssén PE, Unger J, Sintorn IM, editors, Image Analysis - 21st Scandinavian Conference, SCIA 2019, Proceedings. Springer. 2019. p. 28-40. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-20205-7_3

CubiCasa5K: A Dataset and an Improved Multi-task Model for Floorplan Image Analysis

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this