Feature Paper Article
1 Institute of Semiconductors, Chinese Academy of Sciences, Beijing 100083, China
2 Cognitive Computing Technology Joint Laboratory, Wave Group, Beijing 10008, China
3 Graduate School, Northern Arizona University, Flagstaff, AZ 86011, USA
Received 1970-01-01; Accepted 1970-01-01; Issue published 1970-01-01
The Principle of Homology Continuity (PHC) based covering learning method is an effective method to solve the pattern recognition problem. However, PHC and the existence of optimal coverage are not mathematical proven. To address this issue, we firstly give the mathematical description and theoretical proof of PHC. On this basis, the theoretical definition of optimal coverage is introduced. Optimal coverage can determine the internal connections among samples as prior knowledge and use covering neurons to learn prior knowledge. Finally, we propose a kind of covering neuron model, and the effectiveness of which is demonstrated through extensive experiments conducted on the CIFAR-10, LFW, and YTF datasets.
Covering neuron models
Deep learning
Homology continuity
Optimal coverage
Cite This Article
N. Xin, W. Yuebao, T. Weijuan, L. Liang, C. Weiwei, and H. Zhenghua, “A Biomimetic Covering Learning Method Based on Principle of Homology Continuity”, Issue 1, vol. 1, no. 1, pp. 1–13, 2021, doi: TPRIS.2022.7JCBV4VU.

[1]J. Somers, “Is ai riding a one-trick pony?”Technology review, vol. 120, no. 6, pp.29–36, 2017.

[2]Yuille, A. L., & Liu, C. (2021). Deep nets: What have they ever done for vision?.International Journal of Computer Vision, 129(3), 781–802.

[3]Almahairi, A., Ballas, N., Cooijmans, T., Zheng, Y., Larochelle, H., & Courville, A.(2016, June). Dynamic capacity networks. InInternational Conference on MachineLearning(pp. 2549–2558).

[4]Ning, X., Li, W., Tang, B., & He, H. (2018). BULDP: biomimetic uncorrelatedlocality discriminant projection for feature extraction in face recognition.IEEETransactions on Image Processing, 27(5), 2575–2586.

[5]Ning, X., Nan, F., Xu, S., Yu, L., & Zhang, L. (2020). Multi-view frontal face imagegeneration: A survey.Concurrency and Computation: Practice and Experience, e6147.

[6]Ning, X., Li, W., & Liu, W. (2017). A fast single image haze removal method basedon human retina property.IEICE TRANSACTIONS on Information and Systems,100(1), 211–214.

[7]Ning, X., Li, W., & Xu, J. (2018). The principle of homology continuity and geo-metrical covering learning for pattern recognition.International Journal of PatternRecognition and Artificial Intelligence, 32(12), 1850042.

[8]Hu, J., Shen, L., Albanie, S., Sun, G., & Wu, E. (2020). Squeeze-and-Excitation Networks.IEEE Transactions on Pattern Analysis and Machine Intelligence, 42(8),2011–2023.

[9]Woo, S., Park, J., Lee, J. Y., & Kweon, I. S. (2018). Cbam: Convolutional block at-tention module. InProceedings of the European conference on computer vision(ECCV)(pp. 3–19).

[10]Li, X., Wang, W., Hu, X., & Yang, J. (2019). Selective kernel networks. InProceedingsof the IEEE/CVF Conference on Computer Vision and Pattern Recognition(pp. 510–519).

[11]Sabour, S., Frosst, N., & Hinton, G. E. (2017, January). Dynamic Routing BetweenCapsules. InNIPS.

[12]Cheng, X., He, J., He, J., & Xu, H. (2019). Cv-CapsNet: Complex-valued capsulenetwork.IEEE Access, 7, 85492–85499.

[13]Zhu, X., Su, W., Lu, L., Li, B., Wang, X., & Dai, J. (2020). DeformableDETR: Deformable Transformers for End-to-End Object Detection.arXiv preprintarXiv:2010.04159.

[14] Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., Zhai, X., Unterthiner,T., ... & Houlsby, N. (2020). An image is worth 16x16 words: Transformers forimage recognition at scale.arXiv preprintarXiv:2010.11929.

[15]Seung, H. S. (1998). Learning continuous attractors in recurrent networks. InAdvances in neural information processing systems(pp. 654–660).

[16]Seung, H. S., & Lee, D. D. (2000). The manifold ways of perception.science,290(5500), 2268–2269.

[17]Shou-Jue, W. (2002). Bionic (topological) pattern recognition-a new model ofpattern recognition theory and its applications.Acta Electronica Sinica, 10.

[18]Shoujue, W., & Jiangliang, L. (2005). Geometrical learning, descriptive geometry,and biomimetic pattern recognition.Neurocomputing, 67, 9–28.

[19]Zhao, X. R., He, Q., & Shi, Z. Z. (2006, May). Hypersurface classifiers ensemblefor high dimensional data sets. InInternational Symposium on Neural Networks(pp.1299–1304). Springer, Berlin, Heidelberg.

[20]Kamgar-Parsi, B., Lawson, W., & Kamgar-Parsi, B. (2011). Toward developmentof a face recognition system for watchlist surveillance.IEEE Transactions on PatternAnalysis and Machine Intelligence, 33(10), 1925–1937.

[21]Muraina, I. D., Alobaedy, M. M., & Ibrahim, H. H. (2017). A framework forpreserving data integrity during mobile device forensic in open source softwareenvironment. InProceedings of the Free and Open Source Software Conference (FOSSC),Muscat, Oman (pp. 22–26).

[22]Hinton, G. E., Sabour, S., & Frosst, N. (2018, February). Matrix capsules with EMrouting. InInternational conference on learning representations.

[23]Schroff, F., Kalenichenko, D., & Philbin, J. (2015). Facenet: A unified embeddingfor face recognition and clustering. InProceedings of the IEEE conference on computervision and pattern recognition(pp. 815–823).

[24]Deng, J., Guo, J., Xue, N., & Zafeiriou, S. (2019). Arcface: Additive angularmargin loss for deep face recognition. InProceedings of the IEEE/CVF Conferenceon Computer Vision and Pattern Recognition(pp. 4690–4699).

[25]Huang, G. B., Mattar, M., Berg, T., & Learned-Miller, E. (2008, October). La-beled faces in the wild: A database forstudying face recognition in unconstrainedenvironments. InWorkshop on faces in’Real-Life’Images: detection, alignment, andrecognition.

[26]Wolf, L., Hassner, T., & Maoz, I. (2011, June). Face recognition in unconstrainedvideos with matched background similarity. InCVPR 2011(pp. 529–534). IEEE.

[27]Taigman, Y., Yang, M., Ranzato, M. A., & Wolf, L. (2014). Deepface: Closing thegap to human-level performance in face verification. InProceedings of the IEEEconference on computer vision and pattern recognition(pp. 1701–1708).

[28]Wen, Y., Zhang, K., Li, Z., & Qiao, Y. (2016, October). A discriminative feature learning approach for deep face recognition. InEuropean conference on computervision(pp. 499–515).

[29]Liu, W., Wen, Y., Yu, Z., Li, M., Raj, B., & Song, L. (2017). Sphereface: Deephypersphere embedding for face recognition. InProceedings of the IEEE conferenceon computer vision and pattern recognition(pp. 212–220).

[30]Wang, H., Wang, Y., Zhou, Z., Ji, X., Gong, D., Zhou, J., ... & Liu, W. (2018).Cosface: Large margin cosine loss for deep face recognition. InProceedings of theIEEE conference on computer vision and pattern recognition(pp. 5265–5274).

[31] Shi, Y., Yu, X., Sohn, K., Chandraker, M., & Jain, A. K. (2020). Towards universalrepresentation learning for deep face recognition. InProceedings of the IEEE/CVFConference on Computer Vision and Pattern Recognition(pp. 6817–6826).

[32]He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for imagerecognition. InProceedings of the IEEE conference on computer vision and patternrecognition(pp. 770–778).

[33]Carion, N., Massa, F., Synnaeve, G., Usunier, N., Kirillov, A., & Zagoruyko, S. (2020,August). End-to-end object detection with transformers. InEuropean Conferenceon Computer Vision(pp. 213–229).

[34]Zhao, Y., Zeng, Y., & Qiao, G. (2021). Brain-inspired classical conditioning model.Iscience, 24(1), 101980.

[35]Kosiorek, A. R., Sabour, S., Teh, Y. W., Hinton, G. E., & STAFFORD-TOLLEY, M. J.(2019). Stacked capsule autoencoders.Advances in Neural Information ProcessingSystems, 32.