Abstract
In recent years, deep convolutional neural networks (DCNN) have evolved significantly in order to demonstrate remarkable performance in various computer vision tasks. However, their excess storage requirements and heavy computational burden restrict their scope of application, particularly on embedded platforms. This problem has motivated the research community to investigate effective approaches that can reduce computational burden without compromising its performance. Filter pruning is one of the popular ways to reduce the computational burden, where weak or unimportant convolutional filters are eliminated. In this paper, we propose a novel approach for filter pruning based on an adaptive multi-objective particle swarm optimization (AMPSO) to compress and accelerate DCNN. The proposed approach searches for an optimal solution while maintaining the trade-off between network’s performance and computational cost. Extensive experiments on TernausNet and U-Net for high-resolution aerial image segmentation tasks demonstrate the superiority of AMPSO in finding a compact network model.
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11227-023-05150-1/MediaObjects/11227_2023_5150_Fig1_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11227-023-05150-1/MediaObjects/11227_2023_5150_Fig2_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11227-023-05150-1/MediaObjects/11227_2023_5150_Fig3_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11227-023-05150-1/MediaObjects/11227_2023_5150_Fig4_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11227-023-05150-1/MediaObjects/11227_2023_5150_Fig5_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11227-023-05150-1/MediaObjects/11227_2023_5150_Fig6_HTML.png)
Similar content being viewed by others
Availability of data and materials
The datasets used or analyzed during the current study are available in the https://www.airs-dataset.com/, https://project.inria.fr/aerialimagelabeling/.
Notes
References
Ma R, Li S, Zhang B et al (2022) Flexible and generalized real photograph denoising exploiting dual meta attention. IEEE Trans Cybern. https://doi.org/10.1109/TCYB.2022.3170472
Ma R, Li S, Zhang B, Hu H (2022) Meta PID attention network for flexible and efficient real-world noisy image denoising. IEEE Trans Image Process 31:2053–2066. https://doi.org/10.1109/TIP.2022.3150294
Ma R, Zhang B, Zhou Y et al (2022) PID controller-guided attention neural network learning for fast and effective real photographs denoising. IEEE Trans Neural Networks Learn Syst 33:3010–3023. https://doi.org/10.1109/TNNLS.2020.3048031
Ma R, Li S, Zhang B, Li Z (2022) Towards fast and robust real image denoising with attentive neural network and PID controller. IEEE Trans Multimed 24:2366–2377. https://doi.org/10.1109/TMM.2021.3079697
Szegedy C, Liu W, Jia Y et al (2015) Going deeper with convolutions. Proc IEEE Comput Soc Conf Comput Vis Pattern Recognit. https://doi.org/10.1109/CVPR.2015.7298594
He K, Zhang X, Ren S, Sun J (2016) Deep residual learning for image recognition. Proc IEEE Comput Soc Conf Comput Vis Pattern Recognit. https://doi.org/10.1109/CVPR.2016.90
Ma R, Hu H, Xing S, Li Z (2020) Efficient and fast real-world noisy image denoising by combining pyramid neural network and two-pathway unscented Kalman filter. IEEE Trans Image Process 29:3927–3940. https://doi.org/10.1109/TIP.2020.2965294
Ma R, Li S, Zhang B, Li Z (2022) Generative adaptive convolutions for real-world noisy image denoising. Proc AAAI Conf Artif Intell 36:1935–1943. https://doi.org/10.1609/aaai.v36i2.20088
Shi J, Xu J, Tasaka K, Chen Z (2021) SASL: saliency-adaptive sparsity learning for neural network acceleration. IEEE Trans Circuits Syst Video Technol 31:2008–2019. https://doi.org/10.1109/TCSVT.2020.3013170
Xu Y, Fang Y, Peng W, Wu Y (2019) An efficient gaussian sum filter based on prune-cluster-merge scheme. IEEE Access 7:150992–151005. https://doi.org/10.1109/ACCESS.2019.2948139
Sawant SS, Bauer J, Erick FX et al (2022) An optimal-score-based filter pruning for deep convolutional neural networks. Appl Intell. https://doi.org/10.1007/s10489-022-03229-5
Han S, Mao H, Dally WJ (2016) Deep compression: compressing deep neural networks with pruning, trained quantization and Huffman coding. In: 4th International Conference Learn Represent ICLR 2016 - Conference Track Proceedings, pp 1–14
Han S, Pool J, Tran J, Dally WJ (2015) Learning both weights and connections for efficient neural networks. Adv Neural Inf Process Syst 2015:1135–1143
He Y, Kang G, Dong X et al (2018) Soft filter pruning for accelerating deep convolutional neural networks. Int J Conf Artif Intell 2018:2234–2240. https://doi.org/10.24963/ijcai.2018/309
He Y, Dong X, Kang G et al (2020) Asymptotic soft filter pruning for deep convolutional neural networks. IEEE Trans Cybern 50:3594–3604. https://doi.org/10.1109/ICARCV50220.2020.9305319
Jang Y, Lee S, Kim J (2021) Compressing convolutional neural networks by pruning density peak filters. IEEE Access 9:8278–8285. https://doi.org/10.1109/ACCESS.2021.3049470
Lee S, Heo B, Ha JW, Song BC (2020) Filter pruning and re-initialization via latent space clustering. IEEE Access 8:189587–189597. https://doi.org/10.1109/ACCESS.2020.3031031
Chen Z, Chen Z, Lin J et al (2020) Deep neural network acceleration based on low-rank approximated channel pruning. IEEE Trans Circuits Syst I Regul Pap 67:1232–1244. https://doi.org/10.1109/TCSI.2019.2958937
Luo JH, Zhang H, Zhou HY et al (2019) ThiNet: pruning CNN filters for a thinner net. IEEE Trans Pattern Anal Mach Intell 41:2525–2538. https://doi.org/10.1109/TPAMI.2018.2858232
Singh P, Verma VK, Rai P, Namboodiri VP (2020) Acceleration of deep convolutional neural networks using adaptive filter pruning. IEEE J Sel Top Signal Process 14:838–847. https://doi.org/10.1109/JSTSP.2020.2992390
Ayinde BO, Inanc T, Zurada JM (2019) Redundant feature pruning for accelerated inference in deep neural networks. Neural Netw 118:148–158. https://doi.org/10.1016/j.neunet.2019.04.021
Gamanayake C, Jayasinghe L, Ng BKK, Yuen C (2020) Cluster pruning: an efficient filter pruning method for edge AI vision applications. IEEE J Sel Top Signal Process 14(4):802–816. https://doi.org/10.1109/JSTSP.2020.2971418
Sawant SS, Wiedmann M, Göb S et al (2022) Compression of deep convolutional neural network using additional importance-weight-based filter pruning approach. Appl Sci. https://doi.org/10.3390/app122111184
Wang Y, Xu C, Qiu J et al (2018) Towards evolutionary compression. Proc ACM SIGKDD Int Conf Knowl Discov Data Min. https://doi.org/10.1145/3219819.3219970
Tang Y, Ji J, Zhu Y et al (2019) A differential evolution-oriented pruning neural network model for bankruptcy prediction. Complexity. https://doi.org/10.1155/2019/8682124
Wang Z, Li F, Shi G et al (2020) Network pruning using sparse learning and genetic algorithm. Neurocomputing 404:247–256. https://doi.org/10.1016/j.neucom.2020.03.082
Chang J, Lu Y, Xue P, et al (2021) ACP: automatic channel pruning via clustering and swarm intelligence optimization for CNN, pp 1–13
Lin L, Yang Y, Guo Z (2021) AACP: model compression by accurate and automatic channel pruning
Wu T, Li X, Zhou D et al (2021) Differential evolution based layer-wise weight pruning for compressing deep neural networks. Sensors 21:1–20. https://doi.org/10.3390/s21030880
Zhou Y, Yen GG, Yi Z (2021) A knee-guided evolutionary algorithm for compressing deep neural networks. IEEE Trans Cybern 51:1626–1638. https://doi.org/10.1109/TCYB.2019.2928174
Wang B, Sun Y, Xue B, Zhang M (2019) Evolving deep neural networks by multi-objective particle swarm optimization for image classification. Proc Genet Evol Comput Conf. https://doi.org/10.1145/3321707.3321735
Zhou Y, Yen GG, Yi Z (2020) Evolutionary compression of deep neural networks for biomedical image segmentation. IEEE Trans Neural Networks Learn Syst 31:2916–2929. https://doi.org/10.1109/TNNLS.2019.2933879
Zhou Y, Yen GG, Yi Z (2021) Evolutionary shallowing deep neural networks at block levels. IEEE Trans Neural Networks Learn Syst. https://doi.org/10.1109/TNNLS.2021.3059529
Huang J, Sun W, Huang L (2020) Deep neural networks compression learning based on multiobjective evolutionary algorithms. Neurocomputing 378:260–269. https://doi.org/10.1016/j.neucom.2019.10.053
Fernandes FE, Yen GG (2021) Pruning deep convolutional neural networks architectures with evolution strategy. Inf Sci 552:29–47. https://doi.org/10.1016/j.ins.2020.11.009
Deb K, Pratap A, Agarwal S, Meyarivan T (2002) A fast and elitist multiobjective genetic algorithm: NSGA-II. IEEE Trans Evol Comput 6:182–197. https://doi.org/10.1109/4235.996017
Sawant SS, Prabukumar M, Loganathan A et al (2022) Multi-objective multi-verse optimizer based unsupervised band selection for hyperspectral image classification. Int J Remote Sens 43:3990–4024. https://doi.org/10.1080/01431161.2022.2105666
Emmerich MTM, Deutz AH (2018) A tutorial on multiobjective optimization: fundamentals and evolutionary methods. Nat Comput 17:585–609. https://doi.org/10.1007/s11047-018-9685-y
Kennedy J, Eberhart R (1995) Particle swarm optimization. In: Proceedings of ICNN'95 - International Conference on Neural Networks, vol4, Perth, WA, Australia, pp 1942–1948. https://doi.org/10.1109/ICNN.1995.488968.
Liu X, Wu L, Dai C, Chao HC (2021) Compressing CNNs using multi-level filter pruning for the edge nodes of multimedia internet of things. IEEE Internet Things J 4662:1–11. https://doi.org/10.1109/JIOT.2021.3052016
Yu H, Gao Y, Wang J (2020) A multiobjective particle swarm optimization algorithm based on competition mechanism and gaussian variation. Complexity. https://doi.org/10.1155/2020/5980504
Zhu Q, Lin Q, Chen W et al (2017) An external archive-guided multiobjective particle swarm optimization algorithm. IEEE Trans Cybern 47:2794–2808. https://doi.org/10.1109/TCYB.2017.2710133
Ronneberger O, Fischer P, Brox T (2015) U-Net: convolutional networks for biomedical image segmentation medical image computing and computer-assisted intervention (MICCAI). Springer, LNCS, pp 234–241
Iglovikov V, Shvets A (2018) TernausNet: U-Net with VGG11 encoder pre-trained on imagenet for image segmentation. arXiv
Maggiori E, Tarabalka Y, Charpiat G, Alliez P (2017) Can semantic labeling methods generalize to any city? the inria aerial image labeling benchmark. Int Geosci Remote Sens Symp. https://doi.org/10.1109/IGARSS.2017.8127684
Chen Q, Wang L, Wu Y et al (2019) Aerial imagery for roof segmentation: A large-scale dataset towards automatic mapping of buildings. ISPRS J Photogramm Remote Sens 147:42–55. https://doi.org/10.1016/j.isprsjprs.2018.11.011
Satopää V, Albrecht J, Irwin D, Raghavan B (2011) Finding a “kneedle” in a haystack: detecting knee points in system behavior. Proc Int Conf Distrib Comput Syst. https://doi.org/10.1109/ICDCSW.2011.20
Li H, Samet H, Kadav A, et al (2017) Pruning filters for efficient convnets. In: 5th International Conference on Learn Represent ICLR 2017 - Conference Track Proccess, pp 1–5
Mittal D, Bhardwaj S, Khapra MM, Ravindran B (2019) Studying the plasticity in deep convolutional neural networks using random pruning. Mach Vis Appl 30:203–216. https://doi.org/10.1007/s00138-018-01001-9
Luo JH, Wu J (2017) An entropy-based pruning method for CNN compression. arXiv
Acknowledgements
The authors would like to thank the Fraunhofer Institute for Integrated Circuits (IIS) for providing infrastructure for carrying out this research work and the European Research Consortium for Informatics and Mathematics (ERCIM) for the award of a Research Fellowship.
Funding
No funds, Grants, or other support was received.
Author information
Ethics declarations
Conflict of interest
All authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest or non-financial interest in the subject matter or materials discussed in this manuscript.
Consent for publication
Not applicable.
Consent to participate
Not applicable.
Ethical approval
Not applicable.
Human and animal rights
Not applicable.
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
Sawant, S.S., Erick, F.X., Göb, S. et al. An adaptive binary particle swarm optimization for solving multi-objective convolutional filter pruning problem. J Supercomput 79, 13287–13306 (2023). https://doi.org/10.1007/s11227-023-05150-1
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11227-023-05150-1