A Novel U-Net Model Based on Multi-channel Attention Mechanism for Art Aesthetics Image Automatic Segmentation in Design

Shaojie  Zhang

doi:10.6180/jase.202402_27(2).0007

A Novel U-Net Model Based on Multi-channel Attention Mechanism for Art Aesthetics Image Automatic Segmentation in Design

Computer Science and Information Engineering

Shaojie Zhang^1,2This email address is being protected from spambots. You need JavaScript enabled to view it.

¹Henan Institute of Economics and Trade, Zhengzhou 450000 China

²Henan University, Zhengzhou 450046 China

Received: April 9, 2023
Accepted: May 6, 2023
Publication Date: June 20, 2023

Copyright The Author(s). This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are cited.

Download Citation: ||https://doi.org/10.6180/jase.202402_27(2).0007

Traditional image segmentation is inefficient and easy to be affected by objective factors. In order to enhance the accuracy of image segmentation and optimize the detail segmentation effect of image, the paper presents a novel U-Net model based on multi-channel attention mechanism. The skip connection structure in U-Net model is improved based on attention mechanism. Based on attention gate (AG), an efficient channel attention mechanism based on efficient channel attention (ECA) module is added. The new U-Net is divided into three sub-channel attention modules according to the position of the ECA. Finally, comparative experiments on Berkeley data set proves that the proposed method optimizes the image segmentation effect and presents the details of the segmented image completely, which has good application value.

Keywords: Aesthetics image segmentation; U-Net model; multi-channel attention mechanism; skip connection structure; attention gate

[1] Y. Li, G. Blois, F. Kazemifar, and K. T. Christensen, (2021) “A particle-based image segmentation method for phase separation and interface detection in PIV images of immiscible multiphase flow" Measurement Science and Technology 32(9): 095208. DOI: 10.1088/1361-6501/abf0dc.
[2] S. Yin, Y. Zhang, and S. Karim, (2018) “Large scale remote sensing image segmentation based on fuzzy region competition and Gaussian mixture model" IEEE Access 6: 26069–26080. DOI: 10.1109/ACCESS.2018.2834960.
[3] H. Ahmad, S. K. Kim, J. H. Park, and S. Y. Jung, (2022) “Development of two-phase flow regime map for thermally stimulated flows using deep learning and image segmentation technique" International Journal of Multiphase Flow 146: 103869. DOI: 10.1016/j.ijmultiphaseflow.2021.103869.
[4] Q. Chen, L. Zhao, J. Lu, G. Kuang, N. Wang, and Y. Jiang, (2012) “Modified two-dimensional Otsu image segmentation algorithm and fast realisation" IET Image Processing 6(4): 426–433. DOI: 10.1049/iet-ipr.2010.0078.
[5] Q. Chen, B.-l. Xiong, J. Lu, and G.-y. Kuang, (2010) “Improved two-dimensional Otsu image segmentation method and fast recursive realization" Journal of Electronics and Information Technology 32(5): 1100–1104. DOI: 10.3724/SP.J.1146.2009.00627.
[6] E. Tu, L. Cao, J. Yang, and N. Kasabov, (2014) “A novel graph-based k-means for nonlinear manifold clustering and representative selection" Neurocomputing 143: 109–122. DOI: 10.1016/j.neucom.2014.05.067.
[7] L. Zappella, X. Llado, E. Provenzi, and J. Salvi, (2011) “Enhanced local subspace affinity for feature-based motion segmentation" Pattern Recognition 44(2): 454– 470. DOI: 10.1016/j.patcog.2010.08.015.
[8] Y. Zheng, F. Yang, J. Duan, and J. Kurths, (2021) “Quantifying model uncertainty for the observed nonGaussian data by the Hellinger distance" Communications in Nonlinear Science and Numerical Simulation 96: 105720. DOI: 10.1016/j.cnsns.2021.105720.
[9] Y. Bengio, J.-F. Paiement, P. Vincent, O. Delalleau, N. Roux, and M. Ouimet, (2003) “Out-of-sample extensions for lle, isomap, mds, eigenmaps, and spectral clustering" Advances in neural information processing systems 16:
[10] A. Babaeian, A. Bayestehtashk, and M. Bandarabadi, (2015) “Multiple manifold clustering using curvature constrained path" PloS one 10(9): e0137986. DOI: 10.1371/journal.pone.0137986.
[11] I. Tyuryukanov, J. Quiros-Tortos, M. Naglic, M. Popov, M. A. van der Meijden, and V. Terzija. “A post-processing methodology for robust spectral embedded clustering of power networks”. In: IEEE EUROCON 2017-17th International Conference on Smart Technologies. IEEE. 2017, 805–809. DOI: 10.1109/EUROCON.2017.8011221.
[12] X. Ding and X. Li. “Coastline detection in SAR images using multiscale normalized cut segmentation”. In: 2014 IEEE Geoscience and Remote Sensing Symposium. IEEE. 2014, 4447–4449. DOI: 10.1109/IGARSS.2014.6947478.
[13] V. Dakulagi, (2020) “A new Nystrom approximation based efficient coherent DOA estimator for radar applications" AEU-International Journal of Electronics and Communications 124: 153328. DOI: 10.1016/j.aeue.2020.153328.
[14] Y. Wang, H. Wei, X. Ding, and J. Tao, (2020) “Video background/foreground separation model based on nonconvex rank approximation RPCA and superpixel motion detection" IEEE Access 8: 157493–157503. DOI: 10.1109/ACCESS.2020.3018705.
[15] L. He, (2018) “Ray N Guan Y Zhang H Fast large-scale spectral clustering via explicit feature mapping" IEEE Trans. Cybern 49(3): 1058.
[16] W. Song, L. Wen, L. Gao, and X. Li, (2019) “Unsupervised fault diagnosis method based on iterative multimanifold spectral clustering" IET Collaborative Intelligent Manufacturing 1(2): 48–55. DOI: 10.1049/ietcim.2019.0003.
[17] M. Du, S. Ding, and H. Jia, (2016) “Study on density peaks clustering based on k-nearest neighbors and principal component analysis" Knowledge-Based Systems 99: 135–145. DOI: 10.1016/j.knosys.2016.02.001.
[18] F. Chen, Y. Gao, J. Duan, X. Wang, and Y. Yang. “The research on energy partition of multi-energy complementary park based on the improved SMMC algorithm and AHP”. In: 2018 13th IEEE Conference on Industrial Electronics and Applications (ICIEA). IEEE. 2018, 940–945. DOI: 10.1109/ICIEA.2018.8397847.
[19] X. Yao, R. Zhang, J. Hu, K. Chang, X. Liu, and J. Zhao, (2022) “Combining intrinsic dimension and local tangent space for manifold spectral clustering image segmentation" Soft Computing 26(18): 9557–9572. DOI: 10.1007/s00500-022-06751-3.
[20] W. Huang, Q. Hu, Y. Wang, and M. Jiang, (2021) “Multibody nonrigid structure from motion segmentation based on sparse subspace clustering" International Journal of Digital Multimedia Broadcasting 2021: 1–12. DOI: 10.1155/2021/6686179.
[21] Y. Yuan, Z. Xu, and G. Lu, (2021) “SPEDCCNN: spatial pyramid-oriented encoder-decoder cascade convolution neural network for crop disease leaf segmentation" IEEE Access 9: 14849–14866. DOI: 10.1109/ACCESS.2021.3052769.
[22] A. Goel and A. Majumdar, (2021) “Sparse subspace clustering friendly deep dictionary learning for hyperspectral image classification" IEEE Geoscience and Remote Sensing Letters 19: 1–5. DOI: 10.1109/LGRS.2021.3112603.
[23] L. Teng, H. Li, S. Yin, and Y. Sun, (2019) “Improved krill group-based region growing algorithm for image segmentation" International Journal of Image and Data Fusion 10(4): 327–341. DOI: 10.1080/19479832.2019.1604574.
[24] Y. Han and J. C. Ye, (2018) “Framing U-Net via deep convolutional framelets: Application to sparse-view CT" IEEE transactions on medical imaging 37(6): 1418– 1429. DOI: 10.1109/TMI.2018.2823768.
[25] Z. Zhu, X. Lan, T. Zhao, Y. Guo, P. Kojodjojo, Z. Xu, Z. Liu, S. Liu, H. Wang, X. Sun, et al., (2021) “Identification of 27 abnormalities from multi-lead ECG signals: An ensembled SE_ResNet framework with sign loss function" Physiological Measurement 42(6): 065008. DOI: 10.1088/1361-6579/ac08e6.
[26] S. Yin, L. Meng, J. Liu, et al., (2019) “A new apple segmentation and recognition method based on modified fuzzy C-means and hough transform" Journal of Applied Science and Engineering 22(2): 349–354. DOI: 10.6180/jase.201906_22(2).0016.
[27] J. He and D. Jiang, (2021) “Fully automatic model based on se-resnet for bone age assessment" IEEE Access 9: 62460–62466. DOI: 10.1109/ACCESS.2021.3074713.
[28] X. Fu, B. Fang, M. Zhou, and S. Kwong, (2021) “Active contour driven by adaptively weighted signed pressure force combined with Legendre polynomial for image segmentation" Information Sciences 564: 327–342. DOI: 10.1016/j.ins.2021.02.019.
[29] L. Teng, H. Li, S. Yin, S. Karim, and Y. Sun, (2020) “An active contour model based on hybrid energy and fisher criterion for image segmentation" International Journal of Image and Data Fusion 11(1): 97–112. DOI: 10. 1080/19479832.2019.1649309.