Skip to main content

Evaluating Input Perturbation Methods for Interpreting CNNs and Saliency Map Comparison

Part of the Lecture Notes in Computer Science book series (LNIP,volume 12535)


Input perturbation methods occlude parts of an input to a function and measure the change in the function’s output. Recently, input perturbation methods have been applied to generate and evaluate saliency maps from convolutional neural networks. In practice, neutral baseline images are used for the occlusion, such that the baseline image’s impact on the classification probability is minimal. However, in this paper we show that arguably neutral baseline images still impact the generated saliency maps and their evaluation with input perturbations. We also demonstrate that many choices of hyperparameters lead to the divergence of saliency maps generated by input perturbations. We experimentally reveal inconsistencies among a selection of input perturbation methods and find that they lack robustness for generating saliency maps and for evaluating saliency maps as saliency metrics.


  • Saliency methods
  • Saliency maps
  • Saliency metrics
  • Perturbation methods
  • Baseline image
  • RISE
  • MoRF
  • LeRF

This is a preview of subscription content, access via your institution.

Buying options

USD   29.95
Price excludes VAT (USA)
  • DOI: 10.1007/978-3-030-66415-2_8
  • Chapter length: 15 pages
  • Instant PDF download
  • Readable on all devices
  • Own it forever
  • Exclusive offer for individuals only
  • Tax calculation will be finalised during checkout
USD   109.00
Price excludes VAT (USA)
  • ISBN: 978-3-030-66415-2
  • Instant PDF download
  • Readable on all devices
  • Own it forever
  • Exclusive offer for individuals only
  • Tax calculation will be finalised during checkout
Softcover Book
USD   149.99
Price excludes VAT (USA)
Fig. 1.
Fig. 2.
Fig. 3.
Fig. 4.
Fig. 5.
Fig. 6.
Fig. 7.


  1. Abadi, M., et al.: TensorFlow: Large-Scale Machine Learning on Heterogeneous Systems (2015).

  2. Ancona, M., Ceolini, E., Öztireli, A.C., Gross, M.: A unified view of gradient-based attribution methods for deep neural networks. In: NIPS Workshop on Interpreting, Explaining and Visualizing Deep Learning (2017)

    Google Scholar 

  3. Bach, S., Binder, A., Montavon, G., Klauschen, F., Müller, K.R., Samek, W.: On pixel-wise explanations for non-linear classifier decisions by layer-wise relevance propagation. PLoS ONE 10(7), e0130140 (2015)

    CrossRef  Google Scholar 

  4. Bradski, G.: The OpenCV library. Dr. Dobb’s J. Softw. Tools, article id 2236121 (2000).

  5. Brunke, L., Agrawal, P., George, N.: Supplementary Material: Evaluating Input Perturbation Methods for Interpreting CNNs and Saliency Map Comparison (2020)

    Google Scholar 

  6. Chollet, F., et al.: Keras (2015).

  7. Dabkowski, P., Gal, Y.: Real time image saliency for black box classifiers. In: Guyon, I., et al.: (eds.) Advances in Neural Information Processing Systems 30, pp. 6967–6976. Curran Associates, Inc. (2017)

    Google Scholar 

  8. Deng, J., Dong, W., Socher, R., Li, L.J., Li, K., Fei-Fei, L.: ImageNet: a large-scale hierarchical image database. In: CVPR09 (2009)

    Google Scholar 

  9. Fong, R.C., Vedaldi, A.: Interpretable explanations of black boxes by meaningful perturbation. In: The IEEE International Conference on Computer Vision (ICCV), October 2017

    Google Scholar 

  10. He, K., Zhang, X., Ren, S., Sun, J.: Deep residual learning for image recognition. In: The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 2016

    Google Scholar 

  11. Hooker, S., Erhan, D., Kindermans, P.J., Kim, B.: A benchmark for interpretability methods in deep neural networks. In: Advances in Neural Information Processing Systems 32 (2019)

    Google Scholar 

  12. Petsiuk, V., Das, A., Saenko, K.: RISE: randomized input sampling for explanation of black-box models. In: Proceedings of the British Machine Vision Conference (BMVC) (2018)

    Google Scholar 

  13. Samek, W., Binder, A., Montavon, G., Lapuschkin, S., Müller, K.: Evaluating the visualization of what a deep neural network has learned. IEEE Trans. Neural Netw. Learn. Syst. 28(11), 2660–2673 (2017)

    CrossRef  MathSciNet  Google Scholar 

  14. Selvaraju, R.R., Cogswell, M., Das, A., Vedantam, R., Parikh, D., Batra, D.: Grad-CAM: visual explanations from deep networks via gradient-based localization. In: 2017 IEEE International Conference on Computer Vision (ICCV), pp. 618–626, October 2017

    Google Scholar 

  15. Sundararajan, M., Taly, A., Yan, Q.: Axiomatic attribution for deep networks. In: ICML (2017)

    Google Scholar 

  16. Tomsett, R., Harborne, D., Chakraborty, S., Gurram, P., Preece, A.: Sanity checks for saliency metrics. In: Proceedings of the AAAI Conference on Artificial Intelligence (2020)

    Google Scholar 

  17. Zeiler, M.D., Fergus, R.: Visualizing and understanding convolutional networks. In: Fleet, D., Pajdla, T., Schiele, B., Tuytelaars, T. (eds.) ECCV 2014, Part I. LNCS, vol. 8689, pp. 818–833. Springer, Cham (2014).

    CrossRef  Google Scholar 

Download references

Author information

Authors and Affiliations


Corresponding author

Correspondence to Prateek Agrawal .

Editor information

Editors and Affiliations

1 Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary material 1 (pdf 2270 KB)

Rights and permissions

Reprints and Permissions

Copyright information

© 2020 Springer Nature Switzerland AG

About this paper

Verify currency and authenticity via CrossMark

Cite this paper

Brunke, L., Agrawal, P., George, N. (2020). Evaluating Input Perturbation Methods for Interpreting CNNs and Saliency Map Comparison. In: Bartoli, A., Fusiello, A. (eds) Computer Vision – ECCV 2020 Workshops. ECCV 2020. Lecture Notes in Computer Science(), vol 12535. Springer, Cham.

Download citation

  • DOI:

  • Published:

  • Publisher Name: Springer, Cham

  • Print ISBN: 978-3-030-66414-5

  • Online ISBN: 978-3-030-66415-2

  • eBook Packages: Computer ScienceComputer Science (R0)