Abstract
State-of-the-art deep neural networks have achieved great success as an alternative to topology optimization by eliminating the iterative framework of the optimization process. However, models with strong predicting capabilities require massive data, which can be time-consuming, particularly for high-resolution structures. Transfer learning from pre-trained networks has shown promise in enhancing network performance on new tasks with a smaller amount of data. In this study, a U-net-based deep convolutional encoder–decoder network was developed for predicting high-resolution (256 × 256) optimized structures using transfer learning and fine-tuning for topology optimization. Initially, the VGG16 network pre-trained on ImageNet was employed as the encoder for transfer learning. Subsequently, the decoder was constructed from scratch and the network was trained in two steps. Finally, the results of models employing transfer learning and those trained entirely from scratch were compared across various core parameters, including different initial input iterations, fine-tuning epoch numbers, and dataset sizes. Our findings demonstrate that the utilization of transfer learning from the ImageNet pre-trained VGG16 network as the encoder can improve the final predicting performance and alleviate structural discontinuity issues in some cases while reducing training time.
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00521-023-09308-z/MediaObjects/521_2023_9308_Fig1_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00521-023-09308-z/MediaObjects/521_2023_9308_Fig2_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00521-023-09308-z/MediaObjects/521_2023_9308_Fig3_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00521-023-09308-z/MediaObjects/521_2023_9308_Fig4_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00521-023-09308-z/MediaObjects/521_2023_9308_Fig5_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00521-023-09308-z/MediaObjects/521_2023_9308_Fig6_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00521-023-09308-z/MediaObjects/521_2023_9308_Fig7_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00521-023-09308-z/MediaObjects/521_2023_9308_Fig8_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00521-023-09308-z/MediaObjects/521_2023_9308_Fig9_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00521-023-09308-z/MediaObjects/521_2023_9308_Fig10_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00521-023-09308-z/MediaObjects/521_2023_9308_Fig11_HTML.png)
Similar content being viewed by others
Data and material availability
The datasets generated during and/or analyzed during the current study are available at https://github.com/gorkemcanates/Transfer-Learning-for-TO. All necessary information to obtain the topology optimization data and the neural network models that allow to reproduce the results was provided in this paper. The results described in this paper can be replicated by implementing this information.
Code availability
An open-source 2D topology optimization MATLAB code [3] was used to generate data. All codes for neural networks were written in Python 3.7. For neural networks, Pytorch library [33] was used. Our codes are available at https://github.com/gorkemcanates/Transfer-Learning-for-TO
Abbreviations
- θ :
-
Angle of force
- Β :
-
Constants in the optimizer
- \(\mu\) :
-
Mean vector
- \(\sigma\) :
-
Standard deviation vector
- N c :
-
Fixed boundary condition
- N f :
-
Single force application node
- U :
-
Uniform distribution
- v f :
-
Volume fraction
- X min :
-
Maximum of x
- X min :
-
Minimum of x
- x n :
-
Normalized x vector
- \({\widehat{y}}_{i}\) :
-
Predicted value
- \({y}_{i}\) :
-
Target value
- BCE:
-
Binary cross-entropy
- CNN:
-
Convolutional neural network
- DBN:
-
Deep belief network
- FN:
-
False-negative
- FP:
-
False positive
- GAN:
-
Generative adversarial network
- GPU:
-
Graphic processing unit
- MAE:
-
Mean absolute error
- MAPE:
-
Mean absolute percentage error
- MSE:
-
Mean square error
- ReLu:
-
Rectified linear unit
- TN:
-
True negative
- TP:
-
True positive
References
Abdalla A, Cen H, Wan L et al (2019) Fine-tuning convolutional neural network with transfer learning for semantic segmentation of ground-level oilseed rape images in a field with high weed pressure. Comput Electron Agric 167:105091. https://doi.org/10.1016/j.compag.2019.105091
Abueidda DW, Koric S, Sobh NA (2020) Topology optimization of 2D structures with nonlinearities using deep learning. Comput Struct 237:106283. https://doi.org/10.1016/j.compstruc.2020.106283
Andreassen E, Clausen A, Schevenels M et al (2011) Efficient topology optimization in MATLAB using 88 lines of code. Struct Multidiscip Optim 43:1–16. https://doi.org/10.1007/s00158-010-0594-7
Ates GC, Gorguluarslan RM (2021) Two-stage convolutional encoder-decoder network to improve the performance and reliability of deep learning models for topology optimization. Struct Multidiscip Optim 63:1927–1950. https://doi.org/10.1007/s00158-020-02788-w
Banga S, Gehani H, Bhilare S, et al. (2018) 3D Topology optimization using convolutional neural networks. ar**v preprint ar**v:180807440
Cang R, Xu Y, Chen S et al (2017) Microstructure representation and reconstruction of heterogeneous materials via deep belief network for computational material design. J Mech Des 139:071404. https://doi.org/10.1115/1.4036649
Cui B, Chen X, Lu Y (2020) Semantic segmentation of remote sensing images using transfer learning and deep convolutional neural network with dense connection. IEEE Access 8:116744–116755. https://doi.org/10.1109/ACCESS.2020.3003914
Doi S, Sasaki H, Igarashi H (2019) Multi-objective topology optimization of rotating machines using deep learning. IEEE Trans Magn 55:7202605. https://doi.org/10.1109/TMAG.2019.2899934
Frid-Adar M, Ben-Cohen A, Amer R, Greenspan H (2018) Improving the segmentation of anatomical structures in chest radiographs using U-net with an imagenet pre-trained encoder. ar**v 1810.02113
Goodfellow IJ, Pouget-Abadie J, Mirza M, et al (2014) Generative adversarial nets. Adv Neural Inf Process Syst. pp 2672–2680
Gorguluarslan RM, Ates GC (2023) Evaluation of deep learning networks for predicting truss topology optimization results. ASME International Mechanical Engineering Congress and Exposition, Proceedings (IMECE) https://doi.org/10.1115/IMECE2022-95870
He K, Girshick R, Dollar P (2019) Rethinking imageNet pre-training. In: Proceedings of the IEEE International Conference on Computer Vision. pp 4917–4926
He K, Zhang X, Ren S, Sun J (2016) Deep residual learning for image recognition. In: Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition. pp 770–778
Hinton GE, Osindero S, Teh YW (2006) A fast learning algorithm for deep belief nets. Neural Comput 18:1527–1554. https://doi.org/10.1162/neco.2006.18.7.1527
Huang G, Liu Z, Van Der Maaten L, Weinberger KQ (2017) Densely connected convolutional networks. In: Proceedings - 30th IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2017. pp 2261–2269
Huh M, Agrawal P, Efros AA (2016) What makes ImageNet good for transfer learning? ar**v 1608.08614
Iglovikov V, Seferbekov S, Buslaev A, Shvets A (2018) TernausNetV2: Fully convolutional network for instance segmentation. In: IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops. pp 228–232
Iglovikov V, Shvets A (2018) TernausNet: U-Net with VGG11 encoder pre-trained on imagenet for image segmentation. ar**v 1801.05746
Ioffe S, Szegedy C (2015) Batch normalization: Accelerating deep network training by reducing internal covariate shift. In: 32nd International Conference on Machine Learning, ICML 2015. pp 448–456
Jiang H, Nie Z, Yeo R, et al (2020) Stressgan: a generative deep learning model for 2D stress distribution prediction. In: Proceedings of the ASME Design Engineering Technical Conference. Virtual, Online, p V11BT11A023
Kallioras NA, Kazakis G, Lagaros ND (2020) Accelerated topology optimization by means of deep learning. Struct Multidiscip Optim 62:1185–1212. https://doi.org/10.1007/s00158-020-02545-z
Kallioras NA, Lagaros ND (2020) DL-Scale: Deep learning for model upgrading in topology optimization. In: Procedia Manufacturing. pp 433–440
Kingma DP, Ba JL (2015) Adam: A method for stochastic optimization. In: 3rd International Conference on Learning Representations, ICLR 2015 - Conference Track Proceedings. p ar**v:1412.6980
Li B, Huang C, Li X et al (2019) Non-iterative structural topology optimization using deep learning. Comput Aided Des 115:172–180. https://doi.org/10.1016/j.cad.2019.05.038
Lin Q, Hong J, Liu Z et al (2018) Investigation into the topology optimization for conductive heat transfer based on deep learning approach. Int Commun Heat Mass Transfer 97:103–109. https://doi.org/10.1016/j.icheatmasstransfer.2018.07.001
Lin Q, Liu Z, Hong J (2019) Method for directly and instantaneously predicting conductive heat transfer topologies by using supervised deep learning. Int Commun Heat Mass Transfer 109:104368. https://doi.org/10.1016/j.icheatmasstransfer.2019.104368
Liu X, Deng Z, Yang Y (2019) Recent progress in semantic image segmentation. Artif Intell Rev. https://doi.org/10.1007/s10462-018-9641-3
Nair V, Hinton GE (2010) Rectified linear units improve Restricted Boltzmann machines. In: ICML 2010 - Proceedings, 27th International Conference on Machine Learning. pp 807–814
Nakamura K, Suzuki Y (2020) Deep learning-based topological optimization for representing a user-specified design area. ar**v 2004.05461
Nie Z, Lin T, Jiang H, Kara LB (2020) Topologygan: topology optimization using generative adversarial networks based on physical fields over the initial domain. J Mech Des 143:031715. https://doi.org/10.1115/1.4049533
Oh S, Jung Y, Kim S et al (2019) Deep generative design: integration of topology optimization and generative models. J Mech Des 141:111405. https://doi.org/10.1115/1.4044229
Pan SJ, Yang Q (2010) A survey on transfer learning. IEEE Trans Knowl Data Eng 22:1345–1359. https://doi.org/10.1109/TKDE.2009.191
Paszke A, Gross S, Massa F et al (2019) PyTorch: an imperative style, high-performance deep learning library. Adv Neural Inf Process Syst 32:8024–8035
Peng C, **ao T, Li Z, et al (2018) MegDet: A large mini-batch object detector. In: Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition. pp 6181–6189
Pondenkandath V, Alberti M, Eichenberger N, et al (2018) Identifying cross-depicted historical motifs. Proceedings of International Conference on Frontiers in Handwriting Recognition, ICFHR 2018-August:333–338. https://doi.org/10.1109/ICFHR-2018.2018.00065
Ramu P, Thananjayan P, Acar E et al (2022) A survey of machine learning techniques in structural and multidisciplinary optimization. Struct Multidiscip Optimiz 65:1–31. https://doi.org/10.1007/S00158-022-03369-9
Ronneberger O, Fischer P, Brox T (2015) U-net: Convolutional networks for biomedical image segmentation. International Conference on Medical image computing and computer-assisted intervention. Springer, Cham, pp 234–241
Russakovsky O, Deng J, Su H et al (2015) ImageNet large scale visual recognition challenge. Int J Comput Vis 115:211–252. https://doi.org/10.1007/s11263-015-0816-y
Sasaki H, Igarashi H (2019) Topology optimization accelerated by deep learning. IEEE Trans Magn 55:7401305. https://doi.org/10.1109/TMAG.2019.2901906
Sharpe C, Wiest T, Wang P, Seepersad CC (2019) A comparative evaluation of supervised machine learning classification techniques for engineering design applications. J Mech Des 141:121404. https://doi.org/10.1115/1.4044524
Simonyan K, Zisserman A (2015) Very deep convolutional networks for large-scale image recognition. In: 3rd International Conference on Learning Representations, ICLR 2015 - Conference Track Proceedings. San Diego, CA, USA, p ar**v:1409.1556
Singh MS, Pondenkandath V, Zhou B, et al (2017) Transforming sensor data to the image domain for deep learning—An application to footstep detection. Proceedings of the International Joint Conference on Neural Networks 2017-May:2665–2672. https://doi.org/10.1109/IJCNN.2017.7966182
Sosnovik I, Oseledets I (2019) Neural networks for topology optimization. Russ J Numer Anal Math Model 34:215–223. https://doi.org/10.1515/rnam-2019-0018
Studer L, Alberti M, Pondenkandath V, et al (2019) A comprehensive study of imagenet pre-training for historical document image analysis. Proceedings of the International Conference on Document Analysis and Recognition, ICDAR 720–725. https://doi.org/10.1109/ICDAR.2019.00120
Voulodimos A, Doulamis N, Doulamis A, Protopapadakis E (2018) Deep learning for computer vision: a brief review. Comput Intell Neurosci 2018
Wang L, Shi D, Zhang B et al (2023) Deep learning driven real time topology optimization based on improved convolutional block attention (Cba-U-Net) model. Eng Anal Bound Elem 147:112–124. https://doi.org/10.1016/J.ENGANABOUND.2022.11.034
Wu Y, Bai Y, Lan Z, Yao S (2023) A structural-similarity conditional gan method to generate real-time topology for shell-infill structures. Int J Comput Methods. https://doi.org/10.1142/S0219876223410074
Wu Y, He K (2020) Group normalization. Int J Comput Vis 128:742–755. https://doi.org/10.1007/s11263-019-01198-w
Wurm M, Stark T, Zhu XX et al (2019) Semantic segmentation of slums in satellite images using transfer learning on fully convolutional neural networks. ISPRS J Photogramm Remote Sens 150:59–69. https://doi.org/10.1016/j.isprsjprs.2019.02.006
**ang C, Wang D, Pan Y et al (2022) Accelerated topology optimization design of 3D structures based on deep learning. Struct Multidiscip Optim 65:1–18. https://doi.org/10.1007/S00158-022-03194-0/TABLES/7
Yang Z, Li X, Brinson LC et al (2018) Microstructural materials design via deep adversarial learning methodology. J Mech Des 140:111416. https://doi.org/10.1115/1.4041371
Yao Z, Cao Y, Zheng S, et al (2021) Cross-iteration batch normalization. In: Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition
Yu Y, Hur T, Jung J, Jang IG (2019) Deep learning for determining a near-optimal topological design without any iteration. Struct Multidiscip Optim 59:787–799. https://doi.org/10.1007/s00158-018-2101-5
Zhang Y, Chen A, Peng B, et al (2019) A deep Convolutional Neural Network for topology optimization with strong generalization ability. ar**v preprint ar**v:190107761
Zhuang F, Qi Z, Duan K et al (2021) A comprehensive survey on transfer learning. Proc IEEE 109:43–76
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
On behalf of all authors, the corresponding author states that there is no conflict of interest.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Ates, G.C., Gorguluarslan, R.M. Convolutional encoder–decoder network using transfer learning for topology optimization. Neural Comput & Applic 36, 4435–4450 (2024). https://doi.org/10.1007/s00521-023-09308-z
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00521-023-09308-z