Building Envelope Object Detection Using YOLO Models

Norhan Bayomi; Mohanned El Kholy; John E. Fernandez; Senem Velipasalar; Tarek Rakha

doi:10.23919/ANNSIM55834.2022.9859463

Building Envelope Object Detection Using YOLO Models

Norhan Bayomi, Mohanned El Kholy, John E. Fernandez, Senem Velipasalar, Tarek Rakha

Research output: Chapter in Book/Entry/Poem › Conference contribution

5 Scopus citations

Abstract

Building performance significantly influences energy use and indoor thermal conditions tied to the quality of living for its occupants. Therefore, information on building envelopes is essential, especially considering that envelopes and windows can impact 50% of energy loads in the United States. However, current retrofits supporting Building Energy Modelling (BEM) tools face multiple barriers, including time consumption and labor intensity due to manual modeling and calibration processes. This paper proposes using Deep Learning (DL) -based object detection algorithms to detect building envelope components, more specificlly doors, and windows, that can be applied to building energy performance analysis, 3D modeling, and assessment of thermal irregularities. We compare four different versions of the state-of-the-art YOLO V5 model to identify which version best suits the goal of detecting these building components. Results show that YOLO V5_X provides the best performance for detection accuracy.

Original language	English (US)
Title of host publication	Proceedings of the 2022 Annual Modeling and Simulation Conference, ANNSIM 2022
Editors	Cristina Ruiz Martin, Niloufar Emami, Maria Julia Blas, Roya Rezaee
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	617-630
Number of pages	14
ISBN (Electronic)	9781713852889
DOIs	https://doi.org/10.23919/ANNSIM55834.2022.9859463
State	Published - 2022
Event	2022 Annual Modeling and Simulation Conference, ANNSIM 2022 - San Diego, United States Duration: Jul 18 2022 → Jul 20 2022

Publication series

Name	Proceedings of the 2022 Annual Modeling and Simulation Conference, ANNSIM 2022

Conference

Conference	2022 Annual Modeling and Simulation Conference, ANNSIM 2022
Country/Territory	United States
City	San Diego
Period	7/18/22 → 7/20/22

Keywords

BEM
artificial neural networks
building performance
deep learning
object detection

ASJC Scopus subject areas

Artificial Intelligence
Modeling and Simulation
Decision Sciences (miscellaneous)

Access to Document

10.23919/ANNSIM55834.2022.9859463

Cite this

Bayomi, N., Kholy, M. E., Fernandez, J. E., Velipasalar, S., & Rakha, T. (2022). Building Envelope Object Detection Using YOLO Models. In C. R. Martin, N. Emami, M. J. Blas, & R. Rezaee (Eds.), Proceedings of the 2022 Annual Modeling and Simulation Conference, ANNSIM 2022 (pp. 617-630). (Proceedings of the 2022 Annual Modeling and Simulation Conference, ANNSIM 2022). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.23919/ANNSIM55834.2022.9859463

Building Envelope Object Detection Using YOLO Models. / Bayomi, Norhan; Kholy, Mohanned El; Fernandez, John E. et al.
Proceedings of the 2022 Annual Modeling and Simulation Conference, ANNSIM 2022. ed. / Cristina Ruiz Martin; Niloufar Emami; Maria Julia Blas; Roya Rezaee. Institute of Electrical and Electronics Engineers Inc., 2022. p. 617-630 (Proceedings of the 2022 Annual Modeling and Simulation Conference, ANNSIM 2022).

Research output: Chapter in Book/Entry/Poem › Conference contribution

Bayomi, N, Kholy, ME, Fernandez, JE, Velipasalar, S & Rakha, T 2022, Building Envelope Object Detection Using YOLO Models. in CR Martin, N Emami, MJ Blas & R Rezaee (eds), Proceedings of the 2022 Annual Modeling and Simulation Conference, ANNSIM 2022. Proceedings of the 2022 Annual Modeling and Simulation Conference, ANNSIM 2022, Institute of Electrical and Electronics Engineers Inc., pp. 617-630, 2022 Annual Modeling and Simulation Conference, ANNSIM 2022, San Diego, United States, 7/18/22. https://doi.org/10.23919/ANNSIM55834.2022.9859463

Bayomi N, Kholy ME, Fernandez JE, Velipasalar S, Rakha T. Building Envelope Object Detection Using YOLO Models. In Martin CR, Emami N, Blas MJ, Rezaee R, editors, Proceedings of the 2022 Annual Modeling and Simulation Conference, ANNSIM 2022. Institute of Electrical and Electronics Engineers Inc. 2022. p. 617-630. (Proceedings of the 2022 Annual Modeling and Simulation Conference, ANNSIM 2022). doi: 10.23919/ANNSIM55834.2022.9859463

Bayomi, Norhan ; Kholy, Mohanned El ; Fernandez, John E. et al. / Building Envelope Object Detection Using YOLO Models. Proceedings of the 2022 Annual Modeling and Simulation Conference, ANNSIM 2022. editor / Cristina Ruiz Martin ; Niloufar Emami ; Maria Julia Blas ; Roya Rezaee. Institute of Electrical and Electronics Engineers Inc., 2022. pp. 617-630 (Proceedings of the 2022 Annual Modeling and Simulation Conference, ANNSIM 2022).

@inproceedings{3d90933f232f462f837bb2516c68dba6,

title = "Building Envelope Object Detection Using YOLO Models",

abstract = "Building performance significantly influences energy use and indoor thermal conditions tied to the quality of living for its occupants. Therefore, information on building envelopes is essential, especially considering that envelopes and windows can impact 50% of energy loads in the United States. However, current retrofits supporting Building Energy Modelling (BEM) tools face multiple barriers, including time consumption and labor intensity due to manual modeling and calibration processes. This paper proposes using Deep Learning (DL) -based object detection algorithms to detect building envelope components, more specificlly doors, and windows, that can be applied to building energy performance analysis, 3D modeling, and assessment of thermal irregularities. We compare four different versions of the state-of-the-art YOLO V5 model to identify which version best suits the goal of detecting these building components. Results show that YOLO V5_X provides the best performance for detection accuracy.",

keywords = "BEM, artificial neural networks, building performance, deep learning, object detection",

author = "Norhan Bayomi and Kholy, {Mohanned El} and Fernandez, {John E.} and Senem Velipasalar and Tarek Rakha",

note = "Publisher Copyright: {\textcopyright} 2022 SCS.; 2022 Annual Modeling and Simulation Conference, ANNSIM 2022 ; Conference date: 18-07-2022 Through 20-07-2022",

year = "2022",

doi = "10.23919/ANNSIM55834.2022.9859463",

language = "English (US)",

series = "Proceedings of the 2022 Annual Modeling and Simulation Conference, ANNSIM 2022",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "617--630",

editor = "Martin, {Cristina Ruiz} and Niloufar Emami and Blas, {Maria Julia} and Roya Rezaee",

booktitle = "Proceedings of the 2022 Annual Modeling and Simulation Conference, ANNSIM 2022",

}

TY - GEN

T1 - Building Envelope Object Detection Using YOLO Models

AU - Bayomi, Norhan

AU - Kholy, Mohanned El

AU - Fernandez, John E.

AU - Velipasalar, Senem

AU - Rakha, Tarek

PY - 2022

Y1 - 2022

N2 - Building performance significantly influences energy use and indoor thermal conditions tied to the quality of living for its occupants. Therefore, information on building envelopes is essential, especially considering that envelopes and windows can impact 50% of energy loads in the United States. However, current retrofits supporting Building Energy Modelling (BEM) tools face multiple barriers, including time consumption and labor intensity due to manual modeling and calibration processes. This paper proposes using Deep Learning (DL) -based object detection algorithms to detect building envelope components, more specificlly doors, and windows, that can be applied to building energy performance analysis, 3D modeling, and assessment of thermal irregularities. We compare four different versions of the state-of-the-art YOLO V5 model to identify which version best suits the goal of detecting these building components. Results show that YOLO V5_X provides the best performance for detection accuracy.

AB - Building performance significantly influences energy use and indoor thermal conditions tied to the quality of living for its occupants. Therefore, information on building envelopes is essential, especially considering that envelopes and windows can impact 50% of energy loads in the United States. However, current retrofits supporting Building Energy Modelling (BEM) tools face multiple barriers, including time consumption and labor intensity due to manual modeling and calibration processes. This paper proposes using Deep Learning (DL) -based object detection algorithms to detect building envelope components, more specificlly doors, and windows, that can be applied to building energy performance analysis, 3D modeling, and assessment of thermal irregularities. We compare four different versions of the state-of-the-art YOLO V5 model to identify which version best suits the goal of detecting these building components. Results show that YOLO V5_X provides the best performance for detection accuracy.

KW - BEM

KW - artificial neural networks

KW - building performance

KW - deep learning

KW - object detection

UR - http://www.scopus.com/inward/record.url?scp=85138108366&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85138108366&partnerID=8YFLogxK

U2 - 10.23919/ANNSIM55834.2022.9859463

DO - 10.23919/ANNSIM55834.2022.9859463

M3 - Conference contribution

AN - SCOPUS:85138108366

T3 - Proceedings of the 2022 Annual Modeling and Simulation Conference, ANNSIM 2022

SP - 617

EP - 630

BT - Proceedings of the 2022 Annual Modeling and Simulation Conference, ANNSIM 2022

A2 - Martin, Cristina Ruiz

A2 - Emami, Niloufar

A2 - Blas, Maria Julia

A2 - Rezaee, Roya

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2022 Annual Modeling and Simulation Conference, ANNSIM 2022

Y2 - 18 July 2022 through 20 July 2022

ER -

Building Envelope Object Detection Using YOLO Models

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this