Parameterising the SA-UNet using a Genetic Algorithm

Mahsa Mahdinejad; Aidan Murphy; Patrick Healy; Conor Ryan

Parameterising the SA-UNet using a Genetic Algorithm

Mahsa Mahdinejad, Aidan Murphy, Patrick Healy, Conor Ryan

Department of Computer Science & Information Systems

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

Deep learning is an excellent way for effectively addressing image processing, and several Neural Networks designs have been explored in this area. The Spatial Attention U-Net architecture, a version of the famous U-Net but which uses DropBlock and an attention block as well as the common U-Net convolutional blocks, is one notable example. Finding the best combination of hyper-parameters is expensive, time consuming and needs expert input. We show the genetic algorithm can be utilized to automatically determine the optimal combination of Spatial Attention U-Net hyper-parameters to train a model to solve a Retinal Blood Vessel Segmentation problem. Our new approach is able to find a model with an accuracy measure of 0.9855, an improvement from our previous experimentation which found a model with accuracy measure of 0.9751. Our new methods exhibit competitive performance with other state-of-the-art Retinal Blood Vessel Segmentation techniques.

Original language	English
Title of host publication	IJCCI 2022 - Proceedings of the 14th International Joint Conference on Computational Intelligence
Editors	Thomas Back, Bas van Stein, Christian Wagner, Jonathan Garibaldi, H. K. Lam, Marie Cottrell, Faiyaz Doctor, Joaquim Filipe, Kevin Warwick, Janusz Kacprzyk
Publisher	Science and Technology Publications, Lda
Pages	97-104
Number of pages	8
ISBN (Electronic)	9789897586118
Publication status	Published - 2022
Event	14th International Joint Conference on Computational Intelligence, IJCCI 2022 - Valletta, Malta Duration: 24 Oct 2022 → 26 Oct 2022

Publication series

Name	ICETE International Conference on E-Business and Telecommunication Networks (International Joint Conference on Computational Intelligence)
Volume	2022-October
ISSN (Print)	2184-2825

Conference

Conference	14th International Joint Conference on Computational Intelligence, IJCCI 2022
Country/Territory	Malta
City	Valletta
Period	24/10/22 → 26/10/22

Keywords

Deep Learning
Evolutionary Algorithm
Image Segmentation

Cite this

Mahdinejad, M., Murphy, A., Healy, P., & Ryan, C. (2022). Parameterising the SA-UNet using a Genetic Algorithm. In T. Back, B. van Stein, C. Wagner, J. Garibaldi, H. K. Lam, M. Cottrell, F. Doctor, J. Filipe, K. Warwick, & J. Kacprzyk (Eds.), IJCCI 2022 - Proceedings of the 14th International Joint Conference on Computational Intelligence (pp. 97-104). (ICETE International Conference on E-Business and Telecommunication Networks (International Joint Conference on Computational Intelligence); Vol. 2022-October). Science and Technology Publications, Lda.

Mahdinejad, Mahsa ; Murphy, Aidan ; Healy, Patrick et al. / Parameterising the SA-UNet using a Genetic Algorithm. IJCCI 2022 - Proceedings of the 14th International Joint Conference on Computational Intelligence. editor / Thomas Back ; Bas van Stein ; Christian Wagner ; Jonathan Garibaldi ; H. K. Lam ; Marie Cottrell ; Faiyaz Doctor ; Joaquim Filipe ; Kevin Warwick ; Janusz Kacprzyk. Science and Technology Publications, Lda, 2022. pp. 97-104 (ICETE International Conference on E-Business and Telecommunication Networks (International Joint Conference on Computational Intelligence)).

@inproceedings{bdfb1fcb528047b895030011c2b0e523,

title = "Parameterising the SA-UNet using a Genetic Algorithm",

abstract = "Deep learning is an excellent way for effectively addressing image processing, and several Neural Networks designs have been explored in this area. The Spatial Attention U-Net architecture, a version of the famous U-Net but which uses DropBlock and an attention block as well as the common U-Net convolutional blocks, is one notable example. Finding the best combination of hyper-parameters is expensive, time consuming and needs expert input. We show the genetic algorithm can be utilized to automatically determine the optimal combination of Spatial Attention U-Net hyper-parameters to train a model to solve a Retinal Blood Vessel Segmentation problem. Our new approach is able to find a model with an accuracy measure of 0.9855, an improvement from our previous experimentation which found a model with accuracy measure of 0.9751. Our new methods exhibit competitive performance with other state-of-the-art Retinal Blood Vessel Segmentation techniques.",

keywords = "Deep Learning, Evolutionary Algorithm, Image Segmentation",

author = "Mahsa Mahdinejad and Aidan Murphy and Patrick Healy and Conor Ryan",

note = "Publisher Copyright: Copyright {\textcopyright} 2022 by SCITEPRESS - Science and Technology Publications, Lda.; 14th International Joint Conference on Computational Intelligence, IJCCI 2022 ; Conference date: 24-10-2022 Through 26-10-2022",

year = "2022",

language = "English",

series = "ICETE International Conference on E-Business and Telecommunication Networks (International Joint Conference on Computational Intelligence)",

publisher = "Science and Technology Publications, Lda",

pages = "97--104",

editor = "Thomas Back and {van Stein}, Bas and Christian Wagner and Jonathan Garibaldi and Lam, {H. K.} and Marie Cottrell and Faiyaz Doctor and Joaquim Filipe and Kevin Warwick and Janusz Kacprzyk",

booktitle = "IJCCI 2022 - Proceedings of the 14th International Joint Conference on Computational Intelligence",

}

Mahdinejad, M, Murphy, A, Healy, P & Ryan, C 2022, Parameterising the SA-UNet using a Genetic Algorithm. in T Back, B van Stein, C Wagner, J Garibaldi, HK Lam, M Cottrell, F Doctor, J Filipe, K Warwick & J Kacprzyk (eds), IJCCI 2022 - Proceedings of the 14th International Joint Conference on Computational Intelligence. ICETE International Conference on E-Business and Telecommunication Networks (International Joint Conference on Computational Intelligence), vol. 2022-October, Science and Technology Publications, Lda, pp. 97-104, 14th International Joint Conference on Computational Intelligence, IJCCI 2022, Valletta, Malta, 24/10/22.

Parameterising the SA-UNet using a Genetic Algorithm. / Mahdinejad, Mahsa; Murphy, Aidan; Healy, Patrick et al.
IJCCI 2022 - Proceedings of the 14th International Joint Conference on Computational Intelligence. ed. / Thomas Back; Bas van Stein; Christian Wagner; Jonathan Garibaldi; H. K. Lam; Marie Cottrell; Faiyaz Doctor; Joaquim Filipe; Kevin Warwick; Janusz Kacprzyk. Science and Technology Publications, Lda, 2022. p. 97-104 (ICETE International Conference on E-Business and Telecommunication Networks (International Joint Conference on Computational Intelligence); Vol. 2022-October).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Parameterising the SA-UNet using a Genetic Algorithm

AU - Mahdinejad, Mahsa

AU - Murphy, Aidan

AU - Healy, Patrick

AU - Ryan, Conor

PY - 2022

Y1 - 2022

N2 - Deep learning is an excellent way for effectively addressing image processing, and several Neural Networks designs have been explored in this area. The Spatial Attention U-Net architecture, a version of the famous U-Net but which uses DropBlock and an attention block as well as the common U-Net convolutional blocks, is one notable example. Finding the best combination of hyper-parameters is expensive, time consuming and needs expert input. We show the genetic algorithm can be utilized to automatically determine the optimal combination of Spatial Attention U-Net hyper-parameters to train a model to solve a Retinal Blood Vessel Segmentation problem. Our new approach is able to find a model with an accuracy measure of 0.9855, an improvement from our previous experimentation which found a model with accuracy measure of 0.9751. Our new methods exhibit competitive performance with other state-of-the-art Retinal Blood Vessel Segmentation techniques.

AB - Deep learning is an excellent way for effectively addressing image processing, and several Neural Networks designs have been explored in this area. The Spatial Attention U-Net architecture, a version of the famous U-Net but which uses DropBlock and an attention block as well as the common U-Net convolutional blocks, is one notable example. Finding the best combination of hyper-parameters is expensive, time consuming and needs expert input. We show the genetic algorithm can be utilized to automatically determine the optimal combination of Spatial Attention U-Net hyper-parameters to train a model to solve a Retinal Blood Vessel Segmentation problem. Our new approach is able to find a model with an accuracy measure of 0.9855, an improvement from our previous experimentation which found a model with accuracy measure of 0.9751. Our new methods exhibit competitive performance with other state-of-the-art Retinal Blood Vessel Segmentation techniques.

KW - Deep Learning

KW - Evolutionary Algorithm

KW - Image Segmentation

UR - http://www.scopus.com/inward/record.url?scp=85146199528&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:85146199528

T3 - ICETE International Conference on E-Business and Telecommunication Networks (International Joint Conference on Computational Intelligence)

SP - 97

EP - 104

BT - IJCCI 2022 - Proceedings of the 14th International Joint Conference on Computational Intelligence

A2 - Back, Thomas

A2 - van Stein, Bas

A2 - Wagner, Christian

A2 - Garibaldi, Jonathan

A2 - Lam, H. K.

A2 - Cottrell, Marie

A2 - Doctor, Faiyaz

A2 - Filipe, Joaquim

A2 - Warwick, Kevin

A2 - Kacprzyk, Janusz

PB - Science and Technology Publications, Lda

T2 - 14th International Joint Conference on Computational Intelligence, IJCCI 2022

Y2 - 24 October 2022 through 26 October 2022

ER -

Mahdinejad M, Murphy A, Healy P , Ryan C. Parameterising the SA-UNet using a Genetic Algorithm. In Back T, van Stein B, Wagner C, Garibaldi J, Lam HK, Cottrell M, Doctor F, Filipe J, Warwick K, Kacprzyk J, editors, IJCCI 2022 - Proceedings of the 14th International Joint Conference on Computational Intelligence. Science and Technology Publications, Lda. 2022. p. 97-104. (ICETE International Conference on E-Business and Telecommunication Networks (International Joint Conference on Computational Intelligence)).

Parameterising the SA-UNet using a Genetic Algorithm

Abstract

Publication series

Conference

Keywords

Other files and links

Fingerprint

Cite this