计算机科学 ›› 2022, Vol. 49 ›› Issue (6A): 418-423.doi: 10.11896/jsjkx.210700210
王建明1, 陈响育1, 杨自忠2, 史晨阳1, 张宇航1, 钱正坤1
WANG Jian-ming1, CHEN Xiang-yu1, YANG Zi-zhong2, SHI Chen-yang1, ZHANG Yu-hang1, QIAN Zheng-kun1
摘要: 深度学习的效果严重依赖于数据的数量与质量,数据量不足将导致模型过拟合。而在实际应用研究中,大量的高质量样本数据往往较难获取,图像数据尤甚。针对上述问题,以ANIMAL-10数据集为基础,设计了一种基于前景目标提取,使用纯色替换原始背景以实现数据增强的方法,并结合传统数据增强方法构造了新数据集,使用AlexNet,Inception-v3,ResNet-50和VGG-16 4个神经网络模型分析不同颜色背景及不同数据增强方法对模型识别精度的影响。实验结果表明:不同颜色背景对模型识别精度无显著影响。以此为基础,采用绿色背景进行后续的数据增强操作,设计了A,B,C,D 4个数据集并对上述4个模型进行了对比实验,实验结果表明模型对识别精度有显著影响而对数据集无显著影响,但对于AlexNet和Inception-v3模型,包含突出前景数据增强数据集的识别精度较原始图像及传统数据增强方式分别提高了3.78%和4.55%,这说明在小数据集下,突出前景的数据增强方法能使模型更容易注意到并学习到图像的关键特征,从而使模型的表现更好,提高模型的识别精度,在实际的工程应用中具有一定的实践价值。
中图分类号:
[1] ZHU J T,YAO G L,ZHANG G X,et al.Survey of Few Shot Learning of Deep Neural Network[J].CEA,2021,57(7):22-33. [2] YAN L C,YOSHUA B,GEOFFREY H.Deep Learning[J].Nature,2015,521(7553):436-444. [3] JORDAN M I,MITCHELL T M.Machine Learning:Trends,Perspectives,and Prospects[J].Science,2015,349(6245):255-260. [4] ALBAWI S,MOHAMMED T A,AL-ZAWI S.Understandingof aConvolutional Neural Network[C]//International Confe-rence on Engineering and Technology(ICET).IEEE,2017:1-6. [5] NOROUZZADEH M S,NGUYEN A,KOSMALA M,et al.Automatically Identifying,Counting,and Describing Wild Animals in Camera-trap Images with Deep Learning[J].Proceedings of the National Academy of Sciences,2018,115(25):E5716-E5725. [6] DIETTERICH T.Overfitting andUndercomputing in MachineLearning[J].ACM Computing Surveys(CSUR),1995,27(3):326-327. [7] BIEHL M,GHOSH A,HAMMER B.Dynamics and Generalization Ability of LVQ Algorithms[J].Journal of Machine Learning Research,2007,8(8):323-360. [8] KRIZHEVSKY A,SUTSKEVER I,HINTON G.ImageNetClassification with Deep Convolutional Neural Networks[J].Communications of the ACM,2017,60(6):84-90. [9] SIMONYAN K,ZISSERMAN A.Very Deep Convolutional Networks for Large-scale Image Recognition[C]//International Conference on Learning Representations(ICLR).2015. [10] RUSSO F.An Image Enhancement Technique Combining Shar-pening and Noise Reduction[J].IEEE Transactions on Instrumentation and Measurement,2002,51(4):824-828. [11] RANA P,CHOPRA V.A Study on Image Enhancement Techniques[J].IJARCCE,2015,4(5):609-611. [12] LIU S K,TANG P,JIN W D.Study on Catenary Dropper and Support Detection Based on Intelligent Data Augmentation and Improved YOLOv3[J].Computer Science,2020,47(11A):178-182. [13] BAO Y X,LU T L,DU Y H,et al.Deepfake Videos Detection Method Based on i_ResNet34 Model and Data Augmentation[J].Computer Science,2021,48(7):77-85. [14] ZHONG Z,ZHENG L,KANG G,et al.Random Erasing Data Augmentation[J].Proceedings of the AAAI Conference on Artificial Intelligence,2017,34(7):13001-13008. [15] CHEN P,LIU S,ZHAO H,et al.GridMask Data Augmentation[J].arXiv:2001.04086,2020. [16] PEREZ L,WANG J.The Effectiveness of Data Augmentation in Image Classification using Deep Learning[J].arXiv:1712.04621,2017. [17] GIROSI F,JONES M,POGGIO T.Regularization Theory andNeural Networks Architectures[J].Neural computation,1995,7(2):219-269. [18] SCHWARTZ E,KARLINSKY L,SHTOK J,et al.Delta-en-coder:An Effective Sample Synthesis Method for Few-shot Object Recognition[C]//Proceedings of the 32nd International Conference on Neural Information Processing Systems(NIPS'18).2018:2850-2860. [19] WANG Y Q,YAO Q M,KWOK J T,et al.Generalizing from a Few Examples:A Survey on Few-shot Learning[J].ACM Computing Surveys,2020,53(3):1-34. [20] YUE Z Q,ZHANG H W,SUN Q R,et al.Interventional Few-Shot Learning[J].arXiv:2009.13000,2020. [21] LU J,GONG P H,YE J P,et al.Learning from Very Few Samples:A Survey[J].arXiv:2009.02653,2020. [22] IOFFE S,SZEGEDY C.Batch normalization:AcceleratingDeep Network Training by Reducing Internal Covariate shift[C]//International Conference on Machine Learning(ICML).2015:448-456. [23] RIBEIRO M T,SINGH S,GUESTRIN C.“Why Should ITrust You?” Explaining the Predictions of any Classifier[C]//Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining.2016:1135-1144. [24] YU Y,WANG L W,ZHANG Y L.Data Enhancement Algorithm Based on the Correlation between Feature Extraction Preference and Background Color[J].Computer Applications,2019,39(11):3172-3177. [25] CORRADO A.Animal Pictures of 10 Different CategoriesTaken from Google Images[EB/OL].(2019-12-13)[2021-07-21].https://www.kaggle.com/alessiocorrado99/animals10/metadata. |
[1] | 饶志双, 贾真, 张凡, 李天瑞. 基于Key-Value关联记忆网络的知识图谱问答方法 Key-Value Relational Memory Networks for Question Answering over Knowledge Graph 计算机科学, 2022, 49(9): 202-207. https://doi.org/10.11896/jsjkx.220300277 |
[2] | 汤凌韬, 王迪, 张鲁飞, 刘盛云. 基于安全多方计算和差分隐私的联邦学习方案 Federated Learning Scheme Based on Secure Multi-party Computation and Differential Privacy 计算机科学, 2022, 49(9): 297-305. https://doi.org/10.11896/jsjkx.210800108 |
[3] | 周乐员, 张剑华, 袁甜甜, 陈胜勇. 多层注意力机制融合的序列到序列中国连续手语识别和翻译 Sequence-to-Sequence Chinese Continuous Sign Language Recognition and Translation with Multi- layer Attention Mechanism Fusion 计算机科学, 2022, 49(9): 155-161. https://doi.org/10.11896/jsjkx.210800026 |
[4] | 徐涌鑫, 赵俊峰, 王亚沙, 谢冰, 杨恺. 时序知识图谱表示学习 Temporal Knowledge Graph Representation Learning 计算机科学, 2022, 49(9): 162-171. https://doi.org/10.11896/jsjkx.220500204 |
[5] | 李宗民, 张玉鹏, 刘玉杰, 李华. 基于可变形图卷积的点云表征学习 Deformable Graph Convolutional Networks Based Point Cloud Representation Learning 计算机科学, 2022, 49(8): 273-278. https://doi.org/10.11896/jsjkx.210900023 |
[6] | 王剑, 彭雨琦, 赵宇斐, 杨健. 基于深度学习的社交网络舆情信息抽取方法综述 Survey of Social Network Public Opinion Information Extraction Based on Deep Learning 计算机科学, 2022, 49(8): 279-293. https://doi.org/10.11896/jsjkx.220300099 |
[7] | 郝志荣, 陈龙, 黄嘉成. 面向文本分类的类别区分式通用对抗攻击方法 Class Discriminative Universal Adversarial Attack for Text Classification 计算机科学, 2022, 49(8): 323-329. https://doi.org/10.11896/jsjkx.220200077 |
[8] | 姜梦函, 李邵梅, 郑洪浩, 张建朋. 基于改进位置编码的谣言检测模型 Rumor Detection Model Based on Improved Position Embedding 计算机科学, 2022, 49(8): 330-335. https://doi.org/10.11896/jsjkx.210600046 |
[9] | 陈泳全, 姜瑛. 基于卷积神经网络的APP用户行为分析方法 Analysis Method of APP User Behavior Based on Convolutional Neural Network 计算机科学, 2022, 49(8): 78-85. https://doi.org/10.11896/jsjkx.210700121 |
[10] | 朱承璋, 黄嘉儿, 肖亚龙, 王晗, 邹北骥. 基于注意力机制的医学影像深度哈希检索算法 Deep Hash Retrieval Algorithm for Medical Images Based on Attention Mechanism 计算机科学, 2022, 49(8): 113-119. https://doi.org/10.11896/jsjkx.210700153 |
[11] | 孙奇, 吉根林, 张杰. 基于非局部注意力生成对抗网络的视频异常事件检测方法 Non-local Attention Based Generative Adversarial Network for Video Abnormal Event Detection 计算机科学, 2022, 49(8): 172-177. https://doi.org/10.11896/jsjkx.210600061 |
[12] | 檀莹莹, 王俊丽, 张超波. 基于图卷积神经网络的文本分类方法研究综述 Review of Text Classification Methods Based on Graph Convolutional Network 计算机科学, 2022, 49(8): 205-216. https://doi.org/10.11896/jsjkx.210800064 |
[13] | 胡艳羽, 赵龙, 董祥军. 一种用于癌症分类的两阶段深度特征选择提取算法 Two-stage Deep Feature Selection Extraction Algorithm for Cancer Classification 计算机科学, 2022, 49(7): 73-78. https://doi.org/10.11896/jsjkx.210500092 |
[14] | 张颖涛, 张杰, 张睿, 张文强. 全局信息引导的真实图像风格迁移 Photorealistic Style Transfer Guided by Global Information 计算机科学, 2022, 49(7): 100-105. https://doi.org/10.11896/jsjkx.210600036 |
[15] | 戴朝霞, 李锦欣, 张向东, 徐旭, 梅林, 张亮. 基于DNGAN的磁共振图像超分辨率重建算法 Super-resolution Reconstruction of MRI Based on DNGAN 计算机科学, 2022, 49(7): 113-119. https://doi.org/10.11896/jsjkx.210600105 |
|