| 64 | 3 | 0 |
 阅读 |
 下载 |
 被引 |
为了获得更加高效、精度更加准确的图像分类结果，文中提出一种基于MobileNet V2和长短期记忆（LSTM）网络的图像分类框架。使用MobileNet V2对图像进行分类；通过长短期记忆网络维护在前一次图像分类过程中遇到的特征状态信息来增强模型的性能。由于注意力机制可以很容易地集成到标准的卷积神经网络（CNN）架构中，能够以最小的计算负载为代价来提高模型精度和预测精度，因此，文中采用注意力机制改进整个MobileNet V2模型，使其达到预期效果。利用CIFAR-10和CIFAR-100数据集进行分类任务测试，验证模型性能。实验结果表明，在CIFAR-10和CIFAR-100数据集上，所提模型与其他CNN架构相比，有效减少了计算复杂度和参数量，能够推动相关技术在多个领域的应用和发展。
Abstract：An image classification framework based on MobileNet V2 and long and short-term memory (LSTM) network is proposed, which aims to obtain more efficient and accurate image classification results. MobileNet V2 is used to classify images,and the LSTM network is used to enhance the performance of the model by maintaining the feature state information encountered in the previous process of image classification. The attention mechanism can be easily integrated into the standard convolutional neural network (CNN) architecture to improve the model accuracy and prediction accuracy at the expense of the minimum computing load. Therefore, this attention mechanism is adopted to improve the whole MobileNet V2 model to achieve the desired effect. The CIFAR-10 and CIFAR-100 datasets are used for classification tasks in order to verify the performance of the model.Experiments show that the proposed model effectively reduces the computational complexity and required parameters compared with other CNN architectures on CIFAR-10 and CIFAR-100 datasets. To sum up, the proposed model can promote the application and development of related technologies in multiple fields.
[1] 张嫱，何家成，孔令军.面向艺术图像分类的频域增强对比学习方法研究[J].无线电通信技术，2025，51（5）：1046-1055.
[2] 张剑飞，唐泰峰.融合Mamba的乳腺病理图像分类模型[J].黑龙江工业学院学报（综合版），2025，25（7）：83-89.
[3] 王燕，谢俊峰，赵毅力，等.应用变换器双分支模块的高光谱影像树种分类深度学习方法[J]. 东北林业大学学报，2025，53（9）：1-12.
[4] 刘勍，吴忠啸，张亚亚，等.基于深度学习的樱桃图像分类检测[J].现代信息科技，2025，9（11）：64-69.
[5] 杜艳秋，李鑫，康辉，等.基于深度学习的木板激光散斑图像分类研究[J].重庆邮电大学学报（自然科学版），2025，37（5）：677-687.
[6] Zhang Ke, Guo Yurong, Wang Xinsheng, et al. Multiple feature reweight DenseNet for image classification [J]. IEEE Access,2019, 7: 9872-9880.
[7] Yeh J, Chan H T, Hsia C H. ResNeXt with cutout for finger vein analysis [C]// 2021 International Symposium on Intelligent Signal Processing and Communication Systems (ISPACS). [S.l.]:IEEE, 2021: 1-2.
[8] Hu Jie, Shen Li, Sun Gang, et al. Squeeze-and-excitation networks [C]// 2018 IEEE Conference on Computer Vision and Pattern Recognition. Salt Lake City, UT, USA: IEEE, 2018: 7132-7141.
[9] Zhang Xiangyu, Zhou Xinyu, Lin Mengxiao, et al. ShuffleNet:An extremely efficient convolutional neural network for mobile devices [C]// 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City, UT, USA: IEEE, 2018:6848-6856.
[10] Tan Mingxing, Chen Bo, Pang Ruoming, et al. MnasNet:Platform-aware neural architecture search for mobile [C]//2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Long Beach, CA, USA: IEEE, 2019:
2815-2823.
[11] Howard A, Sandler M, Chen Bo, et al. Searching for MobileNetV3[C]// 2019 IEEE/CVF International Conference on Computer Vision (ICCV). Seoul, the Republic of Korea: IEEE, 2019:1314-1324.
[12] Han Kai, Wang Yunhe, Tian Qi, et al. GhostNet: More features from cheap operations [C]// 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Seattle,WA, USA: IEEE, 2020: 1577-1586.
[13] 刘怡，李渊，孙搏遥，等.多尺度异构与跨模态注意力的遥感图像分类[J].空间电子技术，2024，21（6）：57-65.
[14] 唐婷，潘新，罗小玲，等.融合ConvLSTM 和多注意力机制网络的高光谱图像分类[J]. 光谱学与光谱分析，2023，43（8）：2608-2616.
[15] Wang Tianxiao. Domain adaptive English aspect word extraction method based on bidirectional long and short-term memory network and multi-head attention mechanism [J].Journal of Applied Science and Engineering, 2025, 28(12):2461-2469.
[16] Gao Liuxin. GNNTA-BiLSTM: Transformer-attention mechanism bidirectional short-term memory network-based graph neural network for English text attribute extraction [J]. Journal of Applied Science and Engineering, 2025, 28(11): 2209-2216.
[17] 颜乾坤.基于对比学习的水下鱼类图像分类研究[D].西宁：青海师范大学，2025.
[18] 蔡玉红，乔丽娜，肖吉吉，等.基于迁移学习的图像分类模型训练及评估研究[J].电子质量，2025（4）：12-17.
[19] 雷志超.基于自监督和样本权重自适应损失的医学图像分类研究[D].成都：电子科技大学，2025.
[20] 龚谊承，孙莹莹，王晓雨，等.基于正交约束与最小重构误差的对比学习降维及图像分类[J].软件工程，2025，28（7）：25-29.
基本信息:
DOI：10.16652/j.issn.1004-373X.2026.19.005
引用信息:
[1]李航1,陈思行1,殷守林1，2,等.基于长短期记忆网络与MobileNet V2的图像分类研究[J],2026,49(19):31-36.DOI:10.16652/j.issn.1004-373X.2026.19.005.
基金信息:
辽宁省科技联合计划项目（2024JH2/102600106）;辽宁省教育厅项目（LJ232510166008）