Abstract
Recently, researchers have raised several questions related to the explainability of deep learning (DL) model predictions. Indeed, the inherent and undeniable risk remains the abandonment of human control and monitoring in favor of these DL models. Thus, many tools that allow humans to verify the agreement between the DL model predictions and ground truth have been explored. These tools fall into the field of explainable artificial intelligence (XAI) which focuses on proposing methods to explain how the AI-based systems generate their decisions. To contribute to this purpose, we propose in this paper a framework, called DocSegExp, for explaining the decisions of DL models applied for historical document image segmentation (HDIS). The proposed framework is evaluated using three XAI attribution methods on two DL architectures that are performed on a large-scale synthetic dataset having 12K pixel-wise annotated images. Besides, we propose an adaptation of four state-of-the-art metrics (that were previously introduced in the context of object classification) in order to evaluate the generated explanations in the case of a pixel-wise segmentation task.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Alber, M.: Software and application patterns for explanation methods. In: Explainable AI: Interpreting, Explaining and Visualizing Deep Learning, pp. 399–433 (2019)
Ancona, M., Ceolini, E., Öztireli, C., Gross, M.: Towards better understanding of gradient-based attribution methods for deep neural networks. arXiv:1711.06104 (2017)
Anders, C., Weber, L., Neumann, D., Samek, W., Müller, K., Lapuschkin, S.: Finding and removing Clever Hans: using explanation methods to debug and improve deep models. Inf. Fusion, 261–295 (2022)
Arras, L., Osman, A., Samek, W.: CLEVR-XAI: a benchmark dataset for the ground truth evaluation of neural network explanations. Inf. Fusion, 14–40 (2022)
Aubry, M.: Deep learning for historical data analysis. In: SUMAC (2021)
Hastie, T., Tibshirani, R.: Generalized additive models. Statis. Sci. 297–310 (1986)
He, K., Zhang, X., Ren, S., Sun, J.: Deep residual learning for image recognition. In: CVPR, pp. 770–778 (2016)
Kim, B., et al.: Interpretability beyond feature attribution: quantitative testing with concept activation vectors (TCAV). In: ICML, pp. 2668–2677 (2018)
Kokhlikyan, N., et al.: Captum: a unified and generic model interpretability library for PyTorch. arXiv:2009.07896 (2020)
Lin, Y., Lee, W., Celik, Z.: What do you see? Evaluation of explainable artificial intelligence (XAI) interpretability through neural backdoors. arXiv:2009.10639 (2020)
Linardatos, P., Papastefanopoulos, V., Kotsiantis, S.: Explainable AI: a review of machine learning interpretability methods. Entropy (2021)
Lombardi, F., Marinai, S.: Deep learning for historical document analysis and recognition - a survey. J. Imaging (2020)
Long, J., Shelhamer, E., Darrell, T.: Fully convolutional networks for semantic segmentation. In: CVPR, pp. 3431–3440 (2015)
Markewich, L., et al.: Segmentation for document layout analysis: not dead yet. Int. J. Doc. Anal. Recogn. (2022)
Mechi, O., Mehri, M., Ingold, R., Amara, N.E.B.: A two-step framework for text line segmentation in historical Arabic and Latin document images. Int. J. Doc. Anal. Recogn. 197–218 (2021)
Mehri, M., Héroux, P., Mullot, R., Moreux, J., Coüasnon, B., Barrett, B.: ICDAR2019 competition on historical book analysis - HBA2019. In: ICDAR, pp. 1488–1493 (2019)
Monnier, T., Aubry, M.: docExtractor: an off-the-shelf historical document element extraction. In: ICFHR, pp. 91–96 (2020)
Montavon, G.: Gradient-based vs. propagation-based explanations: an axiomatic comparison. In: Explainable AI: Interpreting, Explaining and Visualizing Deep Learning, pp. 253–265 (2019)
Muddamsetty, S., Jahromi, M., Ciontos, A., Fenoy, L., Moeslund, T.: Visual explanation of black-box model: similarity difference and uniqueness (SIDU) method. Pattern Recogn. (2022)
Oliveira, S., Seguin, B., Kaplan, F.: dhSegment: a generic deep-learning approach for document segmentation. In: ICFHR, pp. 7–12 (2018)
Poppi, S., Cornia, M., Baraldi, L., Cucchiara, R.: Revisiting the evaluation of class activation mapping for explainability: a novel metric and experimental analysis. In: CVPR, pp. 2299–2304 (2021)
Rauber, P., Fadel, S., Falcao, A., Telea, A.: Visualizing the hidden activity of artificial neural networks. IEEE Trans. Visual. Comput. Graph. 101–110 (2016)
Ribeiro, M., Singh, S., Guestrin, C.: Why should I trust you? Explaining the predictions of any classifier. In: SIGKDD, pp. 1135–1144 (2016)
Ronneberger, O., Fischer, P., Brox, T.: U-net: convolutional networks for biomedical image segmentation. In: Navab, N., Hornegger, J., Wells, W.M., Frangi, A.F. (eds.) MICCAI 2015. LNCS, vol. 9351, pp. 234–241. Springer, Cham (2015). https://doi.org/10.1007/978-3-319-24574-4_28
Samek, W., Montavon, G., Lapuschkin, S., Anders, C., Müller, K.: Explaining deep neural networks and beyond: a review of methods and applications. IEEE, 247–278 (2021)
Schorr, C., Goodarzi, P., Chen, F., Dahmen, T.: Neuroscope: an explainable AI toolbox for semantic segmentation and image classification of convolutional neural nets. Appl. Sci. (2021)
Selvaraju, R., Cogswell, M., Das, A., Vedantam, R., Parikh, D., Batra, D.: Grad-CAM: visual explanations from deep networks via gradient-based localization. In: ICCV, pp. 618–626 (2017)
Shrikumar, A., Greenside, P., Kundaje, A.: Learning important features through propagating activation differences. arXiv:1704.02685 (2017)
Shrikumar, A., Greenside, P., Shcherbina, A., Kundaje, A.: Not just a black box: Learning important features through propagating activation differences. arXiv:1605.01713 (2016)
Simistira, F., Seuret, M., Eichenberger, N., Garz, A., Liwicki, M., Ingold, R.: DIVA-HisDB: a precisely annotated large dataset of challenging Medieval manuscripts. In: ICFHR, pp. 471–476 (2016)
Singh, A., Sengupta, S., Lakshminarayanan, V.: Explainable deep learning models in medical image analysis. J. Imaging (2020)
Sokolova, M., Lapalme, G.: A systematic analysis of performance measures for classification tasks. Inf. Process. Manag. 427–437 (2009)
Yeh, C., Hsieh, C., Suggala, A., Inouye, D., Ravikumar, P.: On the (in) fidelity and sensitivity of explanations. Adv. Neural Information Processing Systems (2019)
Yosinski, J., Clune, J., Nguyen, A., Fuchs, T., Lipson, H.: Understanding neural networks through deep visualization. arXiv:1506.06579 (2015)
Zeiler, M.D., Fergus, R.: Visualizing and understanding convolutional networks. In: Fleet, D., Pajdla, T., Schiele, B., Tuytelaars, T. (eds.) ECCV 2014. LNCS, vol. 8689, pp. 818–833. Springer, Cham (2014). https://doi.org/10.1007/978-3-319-10590-1_53
Zhou, B., Bau, D., Oliva, A., Torralba, A.: Comparing the interpretability of deep networks via network dissection. In: Explainable AI: Interpreting, Explaining and Visualizing Deep Learning, pp. 243–252 (2019)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Brini, I., Mehri, M., Ingold, R., Essoukri Ben Amara, N. (2022). An End-to-End Framework for Evaluating Explainable Deep Models: Application to Historical Document Image Segmentation. In: Nguyen, N.T., Manolopoulos, Y., Chbeir, R., Kozierkiewicz, A., Trawiński, B. (eds) Computational Collective Intelligence. ICCCI 2022. Lecture Notes in Computer Science(), vol 13501. Springer, Cham. https://doi.org/10.1007/978-3-031-16014-1_10
Download citation
DOI: https://doi.org/10.1007/978-3-031-16014-1_10
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-16013-4
Online ISBN: 978-3-031-16014-1
eBook Packages: Computer ScienceComputer Science (R0)