Skip to main content
SpringerLink
Log in

Clothing classification using transfer learning with squeeze and excitation block

  • Published:
Multimedia Tools and Applications Aims and scope Submit manuscript

Abstract

Automatic clothing classification with deep learning models becomes an inevitable trend in the era of artificial intelligence. Clothing classification methods using conventional convolution neural networks are computationally expensive due to the tremendous training load. To reduce the training time and to improve test accuracy, in this paper, we design a novel transfer learning model and a new training mode. Our proposed transfer learning model is constructed by a pre-trained neural network and novel classification layers with Squeeze-and-Excitation Blocks. The feature maps extracted by the pre-trained CNN are reweighted to enhance the feature selection ability in the blocks. Instead of fine-tuning the un-frozen convolution layers that require extra convolution computing, we separate the classification layers from convolution layers and train it using the adaptive learning rate optimization. Moreover, to facilitate training dataset construction with specific tasks, a novel dataset construction technique is proposed in this paper. Experiments validated that our method can reduce the training time from 25531s to 1419s and also reached classification accuracy of 97.205% that outperformed the existing state-of-art transfer learning methods. It demonstrated that our novel method has potentials in fast data collection, more efficient training for classification models using less training data and simpler computing hardware.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11

Similar content being viewed by others

Notes

  1. image.baidu.com.

References

  1. Ali A, Zhu YM, Zakarya M (2021) A data aggregation based approach to exploit dynamic spatio-temporal correlations for citywide crowd flows prediction in fog computing. Multimed Tools Appl 80(20):31401–31433

    Article  Google Scholar 

  2. Ali A, Zhu YM, Zakarya M (2021) Exploiting dynamic spatio-temporal correlations for citywide traffic flow prediction using attention based neural networks. Inf Sci 577:852–870

    Article  MathSciNet  Google Scholar 

  3. Basly, Hend et al (2021) DTR-HAR: deep temporal residual representation for human activity recognition. The Visual Computer, pp 1–21

  4. Bay H, Ess A, Tuytelaars T, Gool LV (2008) SURF: Speeded up robust features. Comput Vis Image Underst (CVIU) 110(3):346–359

    Article  Google Scholar 

  5. Bayoudh K, Hamdaoui F, Mtibaa A (2020) Transfer learning based hybrid 2D-3D CNN for traffic sign recognition and semantic road detection applied in advanced driver assistance systems. Appl Intell 51(1):124–142

  6. Bucak SS, Jin R, Jain AK (2014) Multiple kernel learning for visual object recognition: a review. IEEE Trans Pattern Anal Mach Intell 36(7):1354–1369

    Article  Google Scholar 

  7. Carreira J, Madeira H, Silva JG (1998) Xception: a technique for the experimental evaluation of dependability in modern computers. IEEE Trans Software Eng 24(2):125–136

    Article  Google Scholar 

  8. Chen J, Liu C (2017) Deep net architectures for visual-based clothing image recognition on large database. Soft Comput 21(11):2923–2939

    Article  Google Scholar 

  9. Chollet F (2017) Xception: Deeplearning with depthwise separable convolutions. Proceedings of the IEEE Conference on Computer Visionand Pattern Recognition, pp 1251–1258

  10. Dalal N, Triggs B (2005) Histograms of oriented gradients for human detection. In: Proceedings of the 2005 IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2005), 1, 886–893

  11. Dimou A, Ataloglou D, Dimitropoulos K, Alvarez F, Daras P (2019)LDS-inspired residual networks. IEEE Trans Circuits Syst Video Technol 29(8):2363–2375

    Article  Google Scholar 

  12. Donati L, Iotti E, Mordonini G, Prati A (2019) Fashion product classification through deep learning and computer vision. Appl Sci 9(7):1385

    Article  Google Scholar 

  13. Ganti V, Sarma AD (2013) Data cleaning: a practical perspective. Morgan & Claypool Publishers

  14. Gu X, Gao F, Tan M, Peng P (2020) Fashion analysis and understanding with artificial intelligence. Inf Process Manage 57(5):1–15

  15. He K, Zhang X, RenS, Sun J (2016) Deep residual learning for image recognition. Proc IEEE Conf Comput Vis Pattern Recognit, pp 770–778

  16. He G, Ji J, Zhang H, Xu Y, Fan J (2020) Feature selection-based hierarchical deep network for image classification. IEEE Access 8:15436–15447

    Article  Google Scholar 

  17. Hidayati SC, You C, Cheng W, Hua K (2018) Learning and recognition of clothing genres from full. IEEE Trans Cybern 48(5):1647–1659

    Article  Google Scholar 

  18. Hu J, Shen L, Sun G, (2018) Squeeze-and-excitation networks. Proceedings of the IEEE Conference on Computer Vision and PatternRecognition (CVPR), pp 7132–7141

  19. Hu J, Song W, Zhang W et al (2019) Deep learning for use in lumber classification tasks. Wood Sci Technol 53:505–517

    Article  Google Scholar 

  20. Khan ZY, Niu Z (2021) CNN with depthwise separable convolutions and combined kernels for rating prediction. Expert Syst Appl 170:114528

  21. Krizhevsky A, Sutskever I, Hinton GE (2012) ImageNet classification with deep convolutional neural networks. Proc Adv Neural Inf Process Syst, pp 1097–1105

  22. Kuang Z, Zhang X, Yu J, Li Z, Fan J (2021) Deep embedding of concept ontology for hierarchical fashion recognition. Neurocomputing 425:191–206

    Article  Google Scholar 

  23. Li R, Lu W, Liang H, Mao Y, Wang X (2018) Multiple features with extreme learning machines for clothing image recognition. IEEE Access 6:36283–36294

    Article  Google Scholar 

  24. Li DS, Cong AR, Guo S (2019) Sewer damage detection from imbalanced CCTV inspection data using deep convolutional neural networks with hierarchical classification. Autom Constr 101:199–208

    Article  Google Scholar 

  25. Lin X, Rivenson Y, Yardimci NT et al (2018)All-optical machine learning using diffractive deep neural networks. Science 361:1004–1008

    Article  MathSciNet  MATH  Google Scholar 

  26. Liu Z, Luo P, Qiu S et al (2016) DeepFashion: Powering robust clothes recognition and retrieval with rich annotations, CVPR

  27. Nguyen N, Bui D, Tran X (2019) A novel hardware architecture for humandetection using HOG-SVM co-optimization. 2019 IEEE Asia Pacific Conference onCircuits and Systems(APCCAS), Bangkok, Thailand, pp 33–36

  28. Pan SJ, Yang Q (2010) A survey on transfer learning. IEEE Trans Knowl Data Eng 22(10):1345–1359

  29. Ravindran P, Costa A, Soares R et al (2018) Classification of CITES-listed and other neotropical Meliaceae wood images using convolutional neural networks. Plant Methods 14:25

    Article  Google Scholar 

  30. Rundo L, Han C, Nagano Y et al (2019) USE-Net_Incorporating Squeeze-and-Excitation blocks into U-Net for prostate zonal segmentation of multi-institutional MRI datasets. Neurocomputing 365:31–43

    Article  Google Scholar 

  31. Srivastava D, Bakthula R, Agarwal S (2019) Image classification using SURF and bag of LBP features constructed by clustering with fixed centers. Multimed Tools Appl 78:14129–14153

    Article  Google Scholar 

  32. Szegedy C, Liu W, Jia YP, et al (2015) Going deeper with convolutions. Proc IEEE Conf Comput Vis Pattern Recognit (CVPR), 1–9

  33. Szegedy C, Vanhoucke V, Ioffe S et al (2016) Rethinking the inception architecture for computer vision. Proc IEEE Conf Comput Vis Pattern Recognit (CVPR), pp 2818–2826

  34. Uzun E, Özhan E, Agun H, Yerlikaya VT, Buluş HN (2020) Automatically discovering relevant images from web pages. IEEE Access 8:208910–208921

    Article  Google Scholar 

  35. Wazarkar S, Keshavamurthy BN (2018) Fashion image classification using matching points with linear convolution. Multimed Tools Appl 77:25941–25958

    Article  Google Scholar 

  36. Weng Y, Wang X, Hua J, Wang H, KangM, Wang F (2019) Forecasting horticultural products price using ARIMA model and neural network based on a large-scale data set collected by web crawler. IEEE Trans Comput Social Syst 6(3):547–553

  37. Xia X, Xu C, Nan B (2017)Inception-v3 for flower classification. 2nd International Conference on Image, Vision and Computing (ICIVC), Chengdu, pp 783–787

  38. Xiang J, Dong T, Pan R, Gao W (2020) Clothing attribute recognition based on RCNN framework using L-Softmax loss. IEEE Access 8:48299–48313

    Article  Google Scholar 

  39. Yamazaki K (2017) A method of classifying crumpled clothing based on image features derived from clothing fabrics and wrinkles. Auton Robot 41(4):865–879

  40. Yian S, Shin K (2019) Hierarchical convolutional neural networks for fashion image classification. Expert Syst Appl 116:328–339

  41. Zhu W, Braun B, Chiang LH, Romagnoli JA (2021) Investigation of transfer learning for image classification and impact on training sample size. Chemom Intell Lab Syst 7639:104269

    Article  Google Scholar 

  42. Zhuang F et al (2021) A comprehensive survey on transfer learning. Proc IEEE 109(1):43–76

    Article  Google Scholar 

  43. Zou M, Zhong Y (2018) Transfer learning for classification of optical satellite image. Sens Imaging 19:6

    Article  Google Scholar 

Download references

Acknowledgements

This work is supported by Xiuying Wang (School of Computer Science, The University of Sydney) who participated in writing and technical editing of the manuscript during the revision.

Author information

Authors and Affiliations

  1. School of Electronic and Information Engineering, Changshu Institute of Technology, Changshu, People’s Republic of China

    Tie-en Xia & Jing-ya Zhang

Authors
  1. Tie-en Xia
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Jing-ya Zhang
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Jing-ya Zhang.

Ethics declarations

Conflict of interest

The authors declare no conflicts of interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Xia, Te., Zhang, Jy. Clothing classification using transfer learning with squeeze and excitation block. Multimed Tools Appl 82, 2839–2856 (2023). https://doi.org/10.1007/s11042-022-13395-w

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11042-022-13395-w

Keywords

Search

Navigation