Substep active deep learning framework for image classification


In image classification, the acquisition of images labels is often expensive and time-consuming. To reduce this labeling cost, active learning is introduced into this field. Although some active learning algorithms have been proposed, they are all single-sampling strategies or combined with multiple-sampling strategies simultaneously (i.e., correlation, uncertainty and label-based measure), without considering the relationship between substep sampling strategies. To this end, we designed a new active learning scheme called substep active deep learning (SADL) for image classification. In SADL, samples were selected by correlation strategy and then determined by the uncertainty and label-based measurement. Finally, it is fed to CNN model training. Experiments were performed with three data sets (i.e., MNIST, Fashion-MNIST and CIFAR-10) to compare against state-of-the-art active learning algorithms, and it can be verified that our substep active deep learning is rational and effective.

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

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5


  1. 1.

    Cheng G, Li Z, Yao X, Guo L, Wei Z (2017) Remote sensing image scene classification using bag of convolutional features. IEEE Geo Sci Remote Sens Lett 14(10):1735–1739

    Article  Google Scholar 

  2. 2.

    Ballester PJ, Richards WG (2007) Ultra fast shape recognition to search compound databases for similar molecular shapes. J Comput Chem 28(10):1711–1723

    Article  Google Scholar 

  3. 3.

    Yang Y, Ramanan D (2011) Articulated pose estimation with flexible mixtures-of–parts. In: Proceedings of the computer vision and pattern recognition, pp 1385–1392

  4. 4.

    Liu X, Fan X, Deng C, Li Z, Su H, Tao D (2016) Multi linear hyper plane hashing. In: The IEEE conference on computer vision and pattern recognition, pp 5119–5127

  5. 5.

    Li IJ, Wu JL, Yeh CH (2017) A fast classification strategy for SVM on the large-scale high-dimensional datasets. Pattern Anal Appl 21(4):1–16

    MathSciNet  Google Scholar 

  6. 6.

    Kanungo T, Mount DM, Netanyahu NS, Piatko CD, Silverman R, Wu AY (2002) An efficient k-means clustering algorithm: analysis and implementation. IEEE Trans Pattern Anal Mach Intell 24(7):881–892

    Article  Google Scholar 

  7. 7.

    Jarraya A, Leray P, Masmoudi A (2014) Discrete exponential Bayesian networks: definition, learning and application for density estimation. Neurocomputing 137:142–149

    Article  Google Scholar 

  8. 8.

    Cheng G, Yang C, Yao X, Guo L, Han J (2018) When deep learning meets metric learning: remote sensing image scene classification via learning discriminative CNNs. IEEE Trans Geosci Remote Sens 56:2811–2821

    Article  Google Scholar 

  9. 9.

    Krizhevsky A, Sutskever I, Hinton GE (2012) Imagenet classification with deep convolutional neural networks. In: Advances in neural information processing systems, pp 1097–1105

  10. 10.

    Käding C, Rodner E, Freytag A, Denzler J (2016) Active and continuous exploration with deep neural networks and expected model output changes. arXiv preprint arXiv:1612.06129

  11. 11.

    Simonyan K, Zisserman A (2014) Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556

  12. 12.

    He K, Zhang X, Ren S, Sun J (2016) Deep residual learning for image recognition. In: International conference on computer vision and pattern recognition

  13. 13.

    Liu D (2018) An active learning algorithm for multi-class classification. Pattern Anal Appl 22:1–13

    Google Scholar 

  14. 14.

    Elahi M, Ricci F, Rubens N (2016) A survey of active learning in collaborative filtering recommender systems. Comput Sci Rev 20:29–50

    MathSciNet  Article  Google Scholar 

  15. 15.

    Lewis D, Gale W (1994) A sequential algorithm for training text classifiers. In: International conference on research and development in information retrieval, pp 3–12

  16. 16.

    Guo Y, Greiner R (2007) Optimistic active learning using mutual information. In: IJCAL, vol 7, pp 823–829

  17. 17.

    Settles B (2010) Active learning literature survey. Univ Wisconsin Madison 52(55–66):11

    Google Scholar 

  18. 18.

    Li X, Yuhong Guo, (2013) Adaptive active learning for image classification. In: International conference on computer vision and pattern recognition, pp 859–866

  19. 19.

    Huang S-j, Jin R, Zhou Z-h (2013) Active learning by querying informative and representative examples. In: Advances in neural information processing systems, pp 1415–1421

  20. 20.

    Li X, Guo Y (2013) Adaptive active learning for image classification. In: Proceedings of CVPR, June, pp 859–866

  21. 21.

    He G, Li Y, Zhao W (2017) An uncertainty and density based active semi-supervised learning scheme for positive unlabeled multivariate time series classification. Knowl Based Syst 124:80–92

    Article  Google Scholar 

  22. 22.

    Angluin D (1988) Queries and concept learning. Mach Learn 2:319–342

    MathSciNet  Google Scholar 

  23. 23.

    Zhao Y, Cao YC, Pan XQ, Xu XN (2009) Tibetan language continuous speech recognition based on active WS-DBN. In: Proceedings of the IEEE international conference on automation and logistics, pp 1558––1562

  24. 24.

    Settles B (2009) Active learning literature survey, computer sciences technical report 1648, University of Wisconsin-Madison

  25. 25.

    Hoi SC, Jin R, Lyu MR (2009) Batch mode active learning with applications to text categorization and image retrieval. IEEE Trans Knowl Data Eng 21:1233–1248

    Article  Google Scholar 

  26. 26.

    Gorriz M, Giro-i Nieto X, Carlier A, Faure E (2017) Cost-effective active learning for melanoma segmentation. In: ML4H: machine learning for health workshop at NIPS 2017, Long Beach, CA, USA

  27. 27.

    Melendez J, van Ginneken B, Maduskar P, Philipsen RHHM, Ayles H, Sanchez CI (2017) On combining multiple-instance learning and active learning for computer-aided detection of tuberculosis. IEEE Trans Med Imaging 35(4):1013–1024

    Article  Google Scholar 

  28. 28.

    Cohn DA, Ghahramani Z, Jordan MI (1994) Improving generalization with active learning. Machine 15(2):201–221

    Google Scholar 

  29. 29.

    Lewis DD, Catlett J (1994) Heterogeneous uncertainty sampling for supervised learning. In: ICML, vol 94, pp 148–156

  30. 30.

    Cohn DA, Ghahramani Z, Jordan MI (1995) Active learning with statistical models. J Artif Intell Res 4:129–145

    Article  Google Scholar 

  31. 31.

    Scheffer T, Decomain C, Wrobel S (2004) Active hidden Markov models for information extraction. In: International symposium on intelligent data analysis. Springer, pp 309–318

  32. 32.

    Shannon CE (2001) A mathematical theory of communication. ACM SIGMOBILE Mob Comput Commun Rev 5(1):3–55

    MathSciNet  Article  Google Scholar 

  33. 33.

    Cohn DA, Ghahramani Z, Jordan MI, Ghahramani Z (1996) Active learning with statistical models. J Artif Intell Res 4(1):129–145

    Article  Google Scholar 

  34. 34.

    Huo L-Z, Tang P (2014) A batch-mode active learning algorithm using region-partitioning diversity for SVM classifier. IEEE J Sel Top Appl Earth Obs Remote Sens 7(4):1036–1046

    Article  Google Scholar 

  35. 35.

    Zhang H, Wang S, Xu X, Chow TWS, Wu QMJ (2018) Tree2vector: learning a vectorial representation for tree-structured data. IEEE Trans Neural Netw Learn Syst 99:1–15

    MathSciNet  Google Scholar 

  36. 36.

    Stark F, Hazırbas C, Triebel R, Cremers D (2015) Captcha recognition with active deep learning. In: Workshop new challenges in neural computation, CiteSeer, p 94

  37. 37.

    Sener O, Savarese S (2018) Active learning for convolutional neural networks: a core-set approach. In: International conference on learning representations

  38. 38.

    Zhou Z, Shin J, Zhang L, Gurudu S, Gotway M, Liang J (2017) Fine-tuning convolutional neural networks for biomedical image analysis: actively and incrementally. In: International conference on computer vision and pattern recognition

  39. 39.

    LeCun Y, Bengio Y, Haffner P (1998) Gradient-based learning applied to document recognition. Process IEEE 86(11):2278–2324

    Article  Google Scholar 

  40. 40.

    Seo Y, Shin K-s (2019) Hierarchical convolutional neural networks for fashion image classification. In: Expert systems with applications vol 116, February, pp 328–339

  41. 41.

    Krizhevsky A, Hinton G (2009) Learning multiple layers of features from tiny images. Technol Rep 11:86–94

    Google Scholar 

  42. 42.

    Wibisono A, Sarwinda D, Mursanto P (2019) Tree stream mining algorithm with Chernoff-bound and standard deviation approach for big data stream. J Big Data 6(1):58

    Article  Google Scholar 

Download references


This project was supported by the National Natural Science Foundation of China (Grant No. 61403331), Program for the Top Young Talents of Higher Learning Institutions of He Bei (Grant No. BJ2017033), Natural Science Foundation of He Bei Province (Grant No. F2016203427) and China Postdoctoral Science Foundation (Grant No. 2015M571280).

Author information



Corresponding author

Correspondence to Ning Gong.

Ethics declarations

Conflict of interest

The authors declare that there is no conflict of interest.

Ethical approval

This article does not contain any studies with human participants performed by any of the authors.

Informed consent

Informed constant was obtained from all individual participants included in the study.

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

Verify currency and authenticity via CrossMark

Cite this article

Li, G., Gong, N. Substep active deep learning framework for image classification. Pattern Anal Applic (2020).

Download citation


  • Convolutional neural network
  • Active learning
  • Substep
  • Image classification