Evaluating saliency map explanations for convolutional neural networks: a user study
The DOI matches its Crossref record; 59 of the 59 references it cites were found in a registry or an index.
Abstract
Convolutional neural networks (CNNs) offer great machine learning performance over a range of applications, but their operation is hard to interpret, even for experts. Various explanation algorithms have been proposed to address this issue, yet limited research effort has been reported concerning their user evaluation. In this paper, we report on an online between-group user study designed to evaluate the performance of "saliency maps" - a popular explanation algorithm for image classification applications of CNNs. Our results indicate that saliency maps produced by the LRP algorithm helped participants to learn about some specific image features the system is sensitive to. However, the maps seem to provide very limited help for participants to anticipate the network's output for new images. Drawing on our findings, we highlight implications for design and further research on explainable AL In particular, we argue the HCI and AI communities should look beyond instance-level explanations.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“Alqaraawi et al. [2] conducted a user study on image classification tasks to evaluate the performance of the saliency map, an XAI technique that highlights input pixels in the original images that contribute to a model prediction and discussed the limited usefulness of saliency map to assist participants to anticipate a model output.”
From Understanding the Effect of Counterfactual Explanations on Trust and Reliance on AI for Human-AI Collaborative Clinical Decision Making (2023) · cites this paper as [2]
In the citing paper
“On the other hand, when individuals are faced with tasks that have lower uncertainty, such as annotation and classifcation task [4, 77, 117], they tend to rely less on the AI advice and rely more on their expertise and judgment.”
From Dealing with Uncertainty: Understanding the Impact of Prognostic Versus Diagnostic Tasks on Trust and Reliance in Human-AI Decision Making (2024) · cites this paper as [4]
In the citing paper
“Unfortunately, the human–XAI interaction techniques face many challenges that need massive work to be solved [103,104].”
From Survey of Explainable AI Techniques in Healthcare (2023) · cites this paper as [104]
In the citing paper
“Unfortunately, the human–XAI interaction techniques face many challenges that need massive work to be solved [103,104].”
From Survey of Explainable AI Techniques in Healthcare (2023) · cites this paper as [104]
In the citing paper
“Such explanations often do not match the end-users' mental models [56], making the explanations ineective [2,40,65] and at times even harmful [13,28,30,47,51].”
From Being Trustworthy is Not Enough: How Untrustworthy Artificial Intelligence (AI) Can Deceive the End-Users and Gain Their Trust (2023) · cites this paper as [2]
1 more claim about this paper
Create a free account to see every claim other papers make about this one.
Citations
Cites
Very Deep Convolutional Networks for Large-Scale Image Recognition
Simonyan, K., & Zisserman, A. (2014) · arXiv · doi:10.48550/arxiv.1409.1556 · Cited by 114,148
ImageNet: A large-scale hierarchical image database
Deng, J., Dong, W., Socher, R., Li, L.-J., Li, K., & Fei-Fei, L. (2009) · 2009 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2009.5206848 · Cited by 65,333
A heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroskedasticity
White, H. (1980) · Econometrica · doi:10.2307/1912934 · Cited by 25,861
Book Review: Pattern Recognition and Machine Learning
Nasrabadi, N. M. (2007) · Journal of Electronic Imaging · doi:10.1117/1.2819119 · Cited by 22,066
"Why Should I Trust You?": Explaining the Predictions of Any Classifier
Ribeiro, M. T., Singh, S., & Guestrin, C. (2016) · Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining · doi:10.1145/2939672.2939778 · Cited by 17,555
Advances in neural information processing systems 7
(1996) · Computers & Mathematics with Applications · doi:10.1016/s0898-1221(96)90463-0 · Cited by 14,281
Pattern Recognition and Machine Learning
(2006) · doi:10.1007/978-0-387-45528-0 · Cited by 9,824
Trust in Automation: Designing for Appropriate Reliance
Lee, J. D., & See, K. A. (2004) · Human Factors: The Journal of the Human Factors and Ergonomics Society · doi:10.1518/hfes.46.1.50_30392 · Cited by 6,605 · cited 2×
“Similarly, the overestimation of a system’s intelligence or capabilities has been shown to impact user interaction negatively [3, 29]. may lead to over-reliance on a system [37], less vigilance towards system failures [61] and unrealistic expectations [61].”
BackgroundPeeking Inside the Black-Box: A Survey on Explainable Artificial Intelligence (XAI)
Adadi, A., & Berrada, M. (2018) · IEEE Access · doi:10.1109/access.2018.2870052 · Cited by 6,294
Lipton, Z. C. (2018) · Queue · doi:10.1145/3236386.3241340 · Cited by 5,427
Research works printed but not linked: 6
- Alex Campolo , Madelyn Sanfilippo , Meredith Whittaker , and Kate Crawford . 2018. AI Now 2017 Report. Microsoft Research (Feb . 2018 ). Alex Campolo, Madelyn Sanfilippo, Meredith Whittaker, and Kate Crawford. 2018. AI Now 2017 Report. Microsoft Research (Feb. 2018).
- 10.1167/7-1.10 · doi:10.1167/7-1.10
- 10.1007/978-3-319-46493--0_1 · doi:10.1007/978-3-319-46493--0_1
- Been Kim , Cynthia Rudin , and Julie A Shah . 2014. The Bayesian Case Model: A Generative Approach for Case-Based Reasoning and Prototype Classification . In Advances in Neural Information Processing Systems 27 , Z. Ghahramani, M. Welling, C. Cortes, N. D. Lawrence, and K. Q. Weinberger (Eds.). Curran Associates, Inc. , 1952 --1960. http://papers.nips.cc/paper/5313-the-bayesian-case-model-a-generative-approach-for-case-based-reasoning-and-prototype-classification.pdf Been Kim, Cynthia Rudin, and Julie A Shah. 2014. The Bayesian Case Model: A Generative Approach for Case-Based Reasoning and Prototype Classification. In Advances in Neural Information Processing Systems 27, Z. Ghahramani, M. Welling, C. Cortes, N. D. Lawrence, and K. Q. Weinberger (Eds.). Curran Associates, Inc., 1952--1960. http://papers.nips.cc/paper/5313-the-bayesian-case-model-a-generative-approach-for-case-based-reasoning-and-prototype-classification.pdf
- Neville Moray . 1999. Mental Models in Theory and Practice. Attention and performance XVII: Cognitive regulation of performance: Interaction of theory and application ( 1999 ), 223--258. Neville Moray. 1999. Mental Models in Theory and Practice. Attention and performance XVII: Cognitive regulation of performance: Interaction of theory and application (1999), 223--258.
- Donald Norman . 2014. On the Relationship between Conceptual and Mental Models . In Gentner et al (e.d) Mental Models . Psychology Press . Donald Norman. 2014. On the Relationship between Conceptual and Mental Models. In Gentner et al (e.d) Mental Models. Psychology Press.
Cited by
Items 201–220 of 220
Developing a Convolutional Neural Network (CNN) Model for Facial Expression Recognition (FER)
Martins, D., & Diesel, L. (2024) · doi:10.63227/373.740.75
Desirability vs. feasibility: a research through design inquiry of explainable AI
Corti, L., & Yang, J. (2025) · Elsevier eBooks · doi:10.1016/b978-0-44-340553-2.00010-1
585Transforming healthcare with machine learning
Mahalakshmi, M. S., Sujatha, S. N. N., Banumathi, S., Purushothaman, V., & Kumar, K. S. (2025) · De Gruyter eBooks · doi:10.1515/9783111503202-013
Visual explanations of neural network predictions
Kumpulainen, J. (2025) · Aaltodoc (Aalto University)
A Test Suite for Saliency Method Evaluation Metrics
Kaspar, M. (2024) · DSpace@MIT (Massachusetts Institute of Technology)
García Ruiz, J. D. (2024) · RUC (Universidade Da Coruña)
Shank, D. B., Canfield, C., & Subramanian, H. V. (2024) · OSF Preprints (OSF Preprints)
Understanding and the Black Box Problem in Artificial Intelligence
Crook, B. (2025) · EPub Bayreuth (University of Bayreuth) · doi:10.15495/epub_ubt_00008992
Attention spatiale artificielle pour des modèles profonds interprétables de qualité embryonnaire
Gomez, T. (2022) · theses.fr (ABES) · doi:10.70675/514b0c84z06afz49a5z8759z40a696a1c6a3 · Cited by 1
Deep learning for the rating of atypical anatomical patterns on MRI data
Hemforth, L. (2024) · theses.fr (ABES) · doi:10.70675/55d681f5z7336z405azb73cze53306570ddc · Cited by 1
From Explanation to Action: A Case Study of Agronomist Workflows
Porfírio, R. P., Santos, P. A., & Madeira, R. N. (2026) · CHI Conference Extended Abstracts on Human Factors in Computing Systems (CHI EA) · doi:10.1145/3772363.3799333
Rollán‐Martínez‐Herrera, M. (2026) · Communities in ADDI (University of the Basque Country)
Trusting the Detector: Comparing Explanations for Deepfake Identification Across Modalities
Kopp, L., Lytovchenko, A., & Cohen, R. A. (2026) · IEEE International Conference on Data Engineering Workshops (ICDEW) · doi:10.1109/icdew71238.2026.00011
Chantzi, V. P., Millington, J., Mariconti, E., & Nakhaeizadeh, S. (2026) · Forensic Science International Synergy · doi:10.1016/j.fsisyn.2026.100712
Gaya-Morey, F. X., Rubio, J. M. B., MacKenzie, I. S., Lacuesta, R., & Manresa-Yee, C. (2025) · Revista de la Asociación Interacción Persona Ordenador (AIPO) · doi:10.65234/interaccion.143
HIL-MDF: A Human-in-the-Loop Multimodal Deepfake Detection System
El-Beheiry, H., Asmah, A., Zaher, O., Abdulwagid, M., & Moussa, S. M. (2026) · IEEE International Conference on Smart Sustainable Systems for Computer and Engineering Applications (3SCEA) · doi:10.1109/3scea68071.2026.11603194
Deviation-Sensitive Black-Box Anomaly Attribution
Idé, T. (2025) · Springer eBooks · doi:10.1007/978-981-96-9108-1_6
Cabitza, F., Gallazzi, E., & Papale, A. (2026) · Proceedings of the ACM on Human-Computer Interaction · doi:10.1145/3816965
Conceptualizing Trustworthiness of AI Systems
Arinaitwe, M., & Hassan, R. (2026) · Communications in computer and information science · doi:10.1007/978-3-032-28454-9_34
Metadata
- Authors
- Alqaraawi, A., Schuessler, M., Weiß, P., Costanza, E., & Berthouze, N.
- Affiliations
- University College London (GB)
- Technische Universität Berlin (DE)
- UCL Interaction Centre, London, United Kingdom
- Venue
- Proceedings of the 25th International Conference on Intelligent User Interfaces
- Publisher
- ACM
- Published
- 2020-03-17
- Pages
- 275-285
- OpenAlex
- W3010300896
- Cited by
- 163
- References
- 53
- Citations by year
- 2026: 9 · 2025: 24 · 2024: 35 · 2023: 34 · 2022: 26 · 2021: 25 · 2020: 10
- FWCI
- 9.06
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- ACM International Conference on Intelligent User Interfaces (IUI) (publishedVersion)
- Funders
- Bundesministerium für Bildung und Forschung (16DII113)
- King Abdulaziz City for Science and Technology
- Engineering and Physical Sciences Research Council (EP/N014243/1)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Explainable Artificial Intelligence (XAI)
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Visual Attention and Saliency Detection
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Keywords
- saliency maps, explainable AI, convolutional neural networks, LRP, model interpretability, image classification, instance-level explanation, feature sensitivity, online user study
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.