Comparative analysis of CNNMVN and MLMVN as frequency domain CNN convolutions
DOI:
https://doi.org/10.17721/1812-5409.2025/1.12Keywords:
convolutional neural networks, CNNMVN, complex-valued neural networks, multi-valued neuron, MLMVN, Image recognition, frequency domainAbstract
Each convolutional layer in any convolutional neural network produces a feature map containing the most important information, which a network needs to recognize respective images. To further improve these neural networks and better understand their capabilities, it is essential to discover, which features are actually extracted and how the images to be recognized are transformed by convolutions resulted from the learning process. This paper presents a comparative analysis of convolutions obtained via two complex-valued neural networks based on multi-valued neurons. The first network is a convolutional neural network based on multi-valued neurons (CNNMVN) which has a traditional convolutional neural network topology except of that it employs complex-valued convolutional kernels in its convolutional part and multi-valued neurons in its fully connected part. The second one is the multi-valued neural network based on multi-valued neurons (MLMVN) which is a fully connected multilayer neural network employed as a convolutional network in the frequency domain.
Considering that both neural networks are complex-valued and the obtained filters operate in the complex domain, the conducted research indicates that the kernels of both networks produce filters similar to existing digital image processing filters. The analysis of CNNMVN kernels revealed that they implement unsharp masking filters and edge detection filters for identifying shapes in images, while the MLMVN kernels enhance specific frequency sub-bands. The latter means that the respective filters are mostly not similar to the ones known as unsharp masking or sharpening filters. Thus, the kernels of both convolutional networks contribute to improving image recognition performance in their own ways.
Pages of the article in the issue: 89 - 96
Language of the article: English
References
Aizenberg, I. (2011). Complex-Valued Neural Networks with Multi-Valued Neurons (Vol. 353). Springer Berlin Heidelberg. https://doi.org/10.1007/978-3-642-20353-4
Aizenberg, I., Herman, J., & Vasko, A. (2022). A Convolutional Neural Network with Multi-Valued Neurons: A Modified Learning Algorithm and Analysis of Performance. 2022 IEEE 13th Annual Ubiquitous Computing, Electronics & Mobile Communication Conference (UEMCON), 0585–0591. https://doi.org/10.1109/UEMCON54665.2022.9965659
Aizenberg, I., & Moraga, C. (2007). Multilayer Feedforward Neural Network Based on Multi-valued Neurons (MLMVN) and a Backpropagation Learning Algorithm. Soft Computing, 11(2), 169–183. https://doi.org/10.1007/s00500-006-0075-5
Aizenberg, I., & Vasko, A. (2020). Convolutional Neural Network with Multi-Valued Neurons. 2020 IEEE Third International Conference on Data Stream Mining & Processing (DSMP), 72–77. https://doi.org/10.1109/DSMP47368.2020.9204076
Aizenberg, I., & Vasko, A. (2023). MLMVN as a Frequency Domain Convolutional Neural Network. 2023 International Conference on Computational Science and Computational Intelligence (CSCI), 341–347. https://doi.org/10.1109/CSCI62032.2023.00061
Aizenberg, I., & Vasko, A. (2024). Frequency-Domain and Spatial-Domain MLMVN-Based Convolutional Neural Networks. Algorithms, 17(8), Article 8. https://doi.org/10.3390/a17080361
Beysolow Ii, T. (2017). Convolutional Neural Networks (CNNs). In T. Beysolow Ii, Introduction to Deep Learning Using R (pp. 101–112). Apress. https://doi.org/10.1007/978-1-4842-2734-3_5
Boonsatit, N., Rajendran, S., Lim, C. P., Jirawattanapanit, A., & Mohandas, P. (2022). New Adaptive Finite-Time Cluster Synchronization of Neutral-Type Complex-Valued Coupled Neural Networks with Mixed Time Delays. Fractal and Fractional, 6(9), 515. https://doi.org/10.3390/fractalfract6090515
Bruna, J., Chintala, S., LeCun, Y., Piantino, S., Szlam, A., & Tygert, M. (2015). A mathematical motivation for complex-valued convolutional networks. https://doi.org/10.48550/ARXIV.1503.03438
Caldeira, M., Martins, P., Cecílio, J., & Furtado, P. (2019). Comparison Study on Convolution Neural Networks (CNNs) vs. Human Visual System (HVS). In S. Kozielski, D. Mrozek, P. Kasprowski, B. Małysiak-Mrozek, & D. Kostrzewa (Eds.), Beyond Databases, Architectures and Structures. Paving the Road to Smart Data Processing and Analysis (Vol. 1018, pp. 111–125). Springer International Publishing. https://doi.org/10.1007/978-3-030-19093-4_9
Chatterjee, S., Tummala, P., Speck, O., & Nürnberger, A. (2023). Complex Network for Complex Problems: A comparative study of CNN and Complex-valued CNN. https://doi.org/10.48550/ARXIV.2302.04584
Fuchs, A., Rock, J., Toth, M., Meissner, P., & Pernkopf, F. (2021). Complex-valued Convolutional Neural Networks for Enhanced Radar Signal Denoising and Interference Mitigation. 2021 IEEE Radar Conference (RadarConf21), 1–6. https://doi.org/10.1109/RadarConf2147009.2021.9455296
Gad, A. F. (2018). Convolutional Neural Networks. In A. F. Gad, Practical Computer Vision Applications Using Deep Learning with CNNs (pp. 183–227). Apress. https://doi.org/10.1007/978-1-4842-4167-7_5
Guberman, N. (2016). On Complex Valued Convolutional Neural Networks (arXiv:1602.09046). arXiv. http://arxiv.org/abs/1602.09046
Hirose, A. (2011). Complex-valued Neural Networks. IEEJ Transactions on Electronics, Information and Systems, 131(1), 2–8. https://doi.org/10.1541/ieejeiss.131.2
Hongo, S., Isokawa, T., Matsui, N., Nishimura, H., & Kamiura, N. (2020). Constructing Convolutional Neural Networks Based on Quaternion. 2020 International Joint Conference on Neural Networks (IJCNN), 1–6. https://doi.org/10.1109/IJCNN48605.2020.9207325
Jarrett, K., Kavukcuoglu, K., Ranzato, M. A., & LeCun, Y. (2009). What is the best multi-stage architecture for object recognition? 2009 IEEE 12th International Conference on Computer Vision, 2146–2153. https://doi.org/10.1109/ICCV.2009.5459469
Kaur, P., & Garg, R. (2020). Towards Convolution Neural Networks (CNNs): A Brief Overview of AI and Deep Learning. In G. Ranganathan, J. Chen, & Á. Rocha (Eds.), Inventive Communication and Computational Technologies (Vol. 89, pp. 399–407). Springer Singapore. https://doi.org/10.1007/978-981-15-0146-3_38
Kobayashi, M. (2017). Symmetric Complex-Valued Hopfield Neural Networks. IEEE Transactions on Neural Networks and Learning Systems, 28(4), 1011–1015. https://doi.org/10.1109/TNNLS.2016.2518672
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2012). ImageNet Classification with Deep Convolutional Neural Networks. Advances in Neural Information Processing Systems, 25. https://doi.org/10.1145/3065386
LeCun, Y., Cortes, C., & Burges, C. J. C. (n.d.). The MNIST Database of handwritten digits. [Dataset]. Retrieved August 9, 2024, from https://www.kaggle.com/datasets/zalando-research/fashionmnist
LeCun, Y., Fu Jie Huang, & Bottou, L. (2004). Learning methods for generic object recognition with invariance to pose and lighting. Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2004. CVPR 2004., 2, 97–104. https://doi.org/10.1109/CVPR.2004.1315150
Lin, L., Liang, L., Jin, L., & Chen, W. (2019). Attribute-Aware Convolutional Neural Networks for Facial Beauty Prediction. Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence, 847–853. https://doi.org/10.24963/ijcai.2019/119
Lin, W., Ding, Y., Wei, H.-L., Pan, X., & Zhang, Y. (2020). LdsConv: Learned Depthwise Separable Convolutions by Group Pruning. Sensors, 20(15), 4349. https://doi.org/10.3390/s20154349
Nitta, T. (2004). Orthogonality of Decision Boundaries in Complex-Valued Neural Networks. Neural Computation, 16(1), 73–97. https://doi.org/10.1162/08997660460734001
Nitta, T. (2008). THE UNIQUENESS THEOREM FOR COMPLEX-VALUED NEURAL NETWORKS WITH THRESHOLD PARAMETERS AND THE REDUNDANCY OF THE PARAMETERS. International Journal of Neural Systems, 18(02), 123–134. https://doi.org/10.1142/S0129065708001439
Popa, C.-A. (2017). Complex-valued convolutional neural networks for real-valued image classification. 2017 International Joint Conference on Neural Networks (IJCNN), 816–822. https://doi.org/10.1109/IJCNN.2017.7965936
Rawat, S., Rana, K. P. S., & Kumar, V. (2021). A novel complex-valued convolutional neural network for medical image denoising. Biomedical Signal Processing and Control, 69, 102859. https://doi.org/10.1016/j.bspc.2021.102859
Sarabu, A., & Santra, A. K. (2021). Human Action Recognition in Videos using Convolution Long Short-Term Memory Network with Spatio-Temporal Networks. Emerging Science Journal, 5(1), 25–33. https://doi.org/10.28991/esj-2021-01254
Sunaga, Y., Natsuaki, R., & Hirose, A. (2020). Similar land-form discovery: Complex absolute-value max pooling in complex-valued convolutional neural networks in interferometric synthetic aperture radar. 2020 International Joint Conference on Neural Networks (IJCNN), 1–7. https://doi.org/10.1109/IJCNN48605.2020.9207122
Suresh, S., Sundararajan, N., & Savitha, R. (2013). Supervised Learning with Complex-valued Neural Networks (Vol. 421). Springer Berlin Heidelberg. https://doi.org/10.1007/978-3-642-29491-4
Valle, M. E. (2014). Complex-Valued Recurrent Correlation Neural Networks. IEEE Transactions on Neural Networks and Learning Systems, 25(9), 1600–1612. https://doi.org/10.1109/TNNLS.2014.2341013
Venkatesan, R., & Li, B. (2017). Modern and Novel Usages of CNNs. In R. Venkatesan & B. Li, Convolutional Neural Networks in Visual Computing (1st ed., pp. 117–146). CRC Press. https://doi.org/10.4324/9781315154282-5
Wu, D., Zhang, J., & Zhao, Q. (2020). A Text Emotion Analysis Method Using the Dual-Channel Convolution Neural Network in Social Networks. Mathematical Problems in Engineering, 2020(1), 6182876. https://doi.org/10.1155/2020/6182876
Xiao, H., Rasul, K., & Vollgraf, R. (2017). Fashion-MNIST: A Novel Image Dataset for Benchmarking Machine Learning Algorithms (Version 2). arXiv. https://doi.org/10.48550/ARXIV.1708.07747
Xiao, L., Zhang, H., Chen, W., Wang, Y., & Jin, Y. (2018). Transformable Convolutional Neural Network for Text Classification. Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence, 4496–4502. https://doi.org/10.24963/ijcai.2018/625
Yadav, S., & Jerripothula, K. R. (2023). FCCNs: Fully Complex-valued Convolutional Networks using Complex-valued Color Model and Loss Function. 2023 IEEE/CVF International Conference on Computer Vision (ICCV), 10655–10664. https://doi.org/10.1109/ICCV51070.2023.00981
Yar, H., Abbas, N., Sadad, T., & Iqbal, S. (2021). Lung Nodule Detection and Classification using 2D and 3D Convolution Neural Networks (CNNs). In L. M. Goyal, T. Saba, A. Rehman, & S. Larabi-Marie-Sainte, Artificial Intelligence and Internet of Things (1st ed., pp. 365–386). CRC Press. https://doi.org/10.1201/9781003097204-17
Zhang, Z., Wang, H., Xu, F., & Jin, Y.-Q. (2017). Complex-Valued Convolutional Neural Network and Its Application in Polarimetric SAR Image Classification. IEEE Transactions on Geoscience and Remote Sensing, 55(12), 7177–7188. https://doi.org/10.1109/TGRS.2017.2743222
Zhang, Z., Wang, Z., Chen, J., & Lin, C. (2022). Complex-Valued Neural Networks Systems with Time Delay: Stability Analysis and (Anti-)Synchronization Control (Vol. 4). Springer Nature Singapore. https://doi.org/10.1007/978-981-19-5450-4
Zhu, X., Xu, Y., Xu, H., & Chen, C. (2018). Quaternion Convolutional Neural Networks. 631–647. https://doi.org/10.48550/arXiv.1903.00658
Downloads
Published
Issue
Section
License
Copyright (c) 2025 Igor Aizenberg, Alexander Vasko
This work is licensed under a Creative Commons Attribution 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).
