计算机科学 ›› 2026, Vol. 53 ›› Issue (7): 24-33.doi: 10.11896/jsjkx.250700003
朱彬, 李晓斌
ZHU Bin, LI Xiaobin
摘要： 针对图像分类任务中图像拓扑特征提取不充分和分类图像结构相似度偏低导致的分类结果准确率低等问题,提出基于注意力机制融合持久同调和卷积神经网络的注意力拓扑卷积融合AETC模型。持久同调方法用于捕捉图像的显著拓扑特征,通过向量化方法提取图像的拓扑特征向量;图像局部特征向量则通过卷积神经网络的卷积池化操作获得。采用注意力得分融合拓扑特征向量和局部特征向量,得到图像的总体特征向量,以解决图像特征提取不充分的问题。使用由可度量的Wasserstein距离和交叉熵损失函数改进得到的Wasserstein交叉熵损失函数约束图像拓扑结构,有效缓解类别间拓扑结构混淆的问题,使AETC模型具有更优的分类性能,从而提高分类鲁棒性与准确性。在3个不同类型的数据集上对采用不同拓扑特征向量化方法的模型进行实验和比较,AETC模型的指标均优于基准模型,准确率ACC值提升2%~11%,AUC值提升1%~7%,F1值提升1%~11%,mAP值提升3%~17%。对于嵌入持续景观向量化方法的经典卷积神经网络模型,其最佳模型ACC值达到95.49%,AUC值达到99.44%,F1值达到95.48%,mAP值达到98.42%。
中图分类号:
| [1]LIU H D,SUN X H,LI Y B,et al.Review of Deep Learning Models for Image Classification Based on Convolutional Neural Networks [J].Chinese Journal of Computer Engineering and Applications,2025,61(11):1-21. [2]GU J X,WANG Z H,KUEN J,et al.Recent Advances in Convolutional Neural Networks [J].Pattern Recognition,2018,77:354-377. [3]LI D,WU H J,WANG Y J,et al.Lightweight Parallel Octave Convolutional Neural Network for Hyperspectral Image Classification[J].Photogrammetric Engineering & Remote Sensing,2023,89(4):233-243. [4]KONG L J,WANG X W,BAO Y C,et al.ASurvey on Medical Image Segmentation Based on Deep Learning[J].Chinese Journal of Radio Communications Technology,2021,47(2):121-130. [5]SUN Y N,XUE B,ZHANG M J,et al.Evolving Deep Convolutional Neural Networks for Image Classification[J].IEEE Transactions on Evolutionary Computation,2020,24(2):394-407. [6]YAN J W,LIN X K,PAN Y Z.Fire image recognition method based on sparse convolutional network pruning [J].Journal of Chongqing Technology and Business University(Natural Science Edition) 2025,42(1):79-84. [7]LECUN Y,BOTTOU L,BENGIO Y,et al.Gradient-based Learning Applied to Document Recognition[J].Proceedings of the IEEE,1998,86(11):2278-2324. [8]SIMONYAN K,ZISSERMAN A.Very Deep Convolutional Networks for Large-scale Image Recognition[C]//Proceedings of the 3rd International Conference on Learning Representations.2015. [9]HE K M,ZHANG X Y,REN S Q,et al.Deep Residual Learning for Image Recognition[C]//Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition.Pisca-taway,NJ:IEEE,2016:770-778. [10]HUANG G,LIU Z,VAN DER MAATEN L,et al.Densely Connected Convolutional Networks[C]//Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition.Piscataway,NJ:IEEE,2017:2261-2269. [11]TAN M X,LE Q V.Efficient Net:Rethinking Model Scaling for Convolutional Neural Networks[C]//36th International Confe-rence on Machine Learning(ICML).2019:10691-10700. [12]CARLSSON G.Topology and Data[J].Bulletin of the American Mathematical Society,2009,46(2):255-308. [13]OUDOT S Y.Persistence Theory:From Quiver Representations to Data Analysis[M].Providence:American Mathematical Society,2015. [14]PUN C S,LEE S X,XIA K.Persistent-homology-based Machine Learning:A Survey and A Comparative Study[J].Artificial Intelligence Review,2022,55(7):5169-5213. [15]ZHANG J L,JU X M.Application of Persistent Homology to Image Classification and Recognition[J].Chinese Journal of Communication on Applied Mathematics and Computation,2017,31(4):494-508. [16]ZHANG Y J.Research on Digital Image Classification Based on Persistent Homology [D].Harbin:Harbin University of Commerce,2023. [17]ZHANG Y P.Plant Leaf Image Classification by Using Persistent Homology and Multi-feature Fusion[D].Wuhan:Wuhan University of Technology,2021. [18]ZHOU Z H.Application of Persistent Homology in Landmark Data's Statistical Analysis[D].Shenyang:Shenyang Normal University,2023. [19]ALI D,ASAAD A,JIMENEZ M J,et al.A Survey of Vectorization Methods in Topological Data Analysis[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2023,45(12):14069-14080. [20]WAGNER H,CHEN C,VUCINI E.Efficient Computation of Persistent Homology for Cubical Data[C]//Topological Me-thods in Data Analysis and Visualization II:Theory,Algorithms,and Applications.2011:91-106. [21]EDELSBRUNNER H,HARER J.Computational Topology:An Introduction[M].American Mathematical Society,2009. [22]周建伟.代数拓扑讲义[M].北京:科学出版社,2007. [23]ALLEN H.Algebraic Topology[M].Cambridge:Cambridge University Press,2001. [24]TURNER K,ROBINS V,MORGAN J.The Extended Persistent Homology Transform of Manifolds with Boundary[J].Journal of Applied and Computational Topology,2024,8(7):2111-2154. [25]SAMVELYAN K,ZHANG J,POLTERO VICH L,et al.Topo-logical Persistence in Geometry and Analysis(University Lecture Series)[M].American Mathematical Society,2020. [26]CRAWLEY-BOEVEY W.Decomposition of Pointwise Finite dimensional Persistence Modules[J].Journal of Algebra and Its Applications,2015,14(5):1550066. [27]STUCKI N,PAETZOLD J C,SHIT S,et al.Topologically Faithful Image Segmentation via Induced Matching of Persis-tence Barcodes[C]//International Conference on Machine Learning.2023:32698-32727. [28]FATEMA S,NUWAGIRA B,CHAKRABORTY S,et al.TopOC:Topological Deep Learning for Ovarian and Breast Cancer Diagnosis[C]//International Workshop on Topology and Graph-Informed Imaging Informatics.2024:22-32. [29]CONTIF,MORONI D,PASCALI M A.A Topological Machine Learning Pipeline for Classification[J].Mathematics,2022,10(17):3086. [30]MURRAY N,PERRONNIN F.Generalized Max Pooling[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2014:2473-2480. [31]VASWANI A,SHAZEER N,PARMAR N,et al.Attention is All You Need[C]//31st Annual Conference on Neural Information Processing Systems(NIPS).2017:5999-6009. [32]HAHNLOSER R H,SARPESHKAR R,MAH OWALD M A,et al.Digital Selection and Analogue Amplification Coexist in A Cortex-inspired Silicon Circuit[J].Nature,2000,405(6789):947-951. [33]PANLAQUI B,DEB D.Convolutional Neural Networks:An Analysis of Varying Factors[C]//IEEE Southeast Conference.2023:348-349. [34]XIAO H,RASUL K,VOLLGRAF R.Fashio-nmnist:A Novel Image Dataset for Benchmarking Machine Learning Algorithms [J].arXiv:1708.07747,2017. [35]KRIZHEVSKY A,HINTON G.Learning Multiple Layers of Features from Tiny Images[J].Handbook of Systemic Autoimmune Diseases,2009,1(4). |
|
||
首页