Explainable GeoAI: can saliency maps help interpret artificial intelligence’s learning process? An empirical study on natural feature detection
Open accessThe DOI matches its Crossref record; 42 of the 45 references it cites were found in a registry or an index, and 1 are not checked yet.
Abstract
Improving the interpretability of geospatial artificial intelligence (GeoAI) models has become critically important to open the ‘black box’ of complex AI models, such as deep learning. This paper compares popular saliency map generation techniques and their strengths and weaknesses in interpreting GeoAI and deep learning models’ reasoning behaviors, particularly when applied to geospatial analysis and image processing tasks. We surveyed two broad classes of model explanation methods: perturbation-based and gradient-based methods. The former identifies important image areas, which help machines make predictions by modifying a localized area of the input image. The latter evaluates the contribution of every single pixel of the input image to the model’s prediction results through gradient backpropagation. In this study, three algorithms—the occlusion method, the integrated gradients method, and the class activation map method—are examined for a natural feature detection task using deep learning. The algorithms’ strengths and weaknesses are discussed, and the consistency between model-learned and human-understandable concepts for object recognition is also compared. The experiments used two GeoAI-ready datasets to demonstrate the generalizability of the research findings.
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
“Through data-driven analysis, deep learning models have demonstrated the capability to recognize distinctive flood patterns and subtle flood boundaries, even in regions under complex and rapidly changing conditions.”
From Assessment of a new GeoAI foundation model for flood inundation mapping (2023) · cites this paper as [13]
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 · cited 2×
“The classification model was trained using a pretrained VGG16 CNN (Simonyan and Zisserman 2014).”
Method"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
Visualizing and understanding convolutional networks
Zeiler, M. D., & Fergus, R. (2014) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-10590-1_53 · Cited by 15,839
Advances in neural information processing systems 7
(1996) · Computers & Mathematics with Applications · doi:10.1016/s0898-1221(96)90463-0 · Cited by 14,281
Advances in Neural Information Processing Systems 29
Teymur, O., Zygalakis, K., & Calderhead, B. (2016) · Cited by 13,887
Learning Deep Features for Discriminative Localization
Zhou, B., Khosla, A., Lapedriza, A., Oliva, A., & Torralba, A. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.319 · Cited by 11,428 · cited 6×
“Compared to methods such as CAM (Zhou et al. 2016), none of the methods studied in this paper require network architecture changes; therefore, there is no need to retrain the deep learning models.”
MethodInfluentialGrad-CAM: Visual Explanations from Deep Networks via Gradient-Based Localization
Selvaraju, R. R., Cogswell, M., Das, A., Vedantam, R., Parikh, D., & Batra, D. (2020) · International Journal of Computer Vision · doi:10.1007/s11263-019-01228-7 · Cited by 6,909 · cited 7×
“In each experiment, we applied three model explanation techniques – including occlusion sensitivity by Zeiler and Fergus (2014), Grad-CAM by Selvaraju et al. (2020), and the integrated gradients method by Sundararajan et al. (2017) – to explain the model’s decision-making process.”
BackgroundMethodInfluentialDeep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps
Simonyan, K., Vedaldi, A., & Zisserman, A. (2013) · arXiv · doi:10.48550/arxiv.1312.6034 · Cited by 5,370 · cited 4×
“Some recent methods calculate gradients at the input space (Simonyan et al. 2014, Zeiler and Fergus 2014, Springenberg et al. 2015), while others need only gradients at the middle layers (Zhou et al. 2016, Selvaraju et al. 2020).”
BackgroundMethodInfluentialAxiomatic attribution for deep networks
Sundararajan, M., Taly, A., & Yan, Q. (2017) · arXiv · doi:10.48550/arxiv.1703.01365 · Cited by 4,087 · cited 7×
“In each experiment, we applied three model explanation techniques – including occlusion sensitivity by Zeiler and Fergus (2014), Grad-CAM by Selvaraju et al. (2020), and the integrated gradients method by Sundararajan et al. (2017) – to explain the model’s decision-making process.”
BackgroundMethodInfluentialGrad-CAM++: Generalized Gradient-Based Visual Explanations for Deep Convolutional Networks
Chattopadhay, A., Sarkar, A., Howlader, P., & Balasubramanian, V. N. (2018) · 2018 IEEE Winter Conference on Applications of Computer Vision (WACV) · doi:10.1109/wacv.2018.00097 · Cited by 3,639 · cited 1×
“Methods that have followed up Grad-CAM include GradCAMþþ (Chattopadhay et al. 2018) for the localization of multi-objects belonging to the same class and Score-CAM (Wang et al. 2020), which obtains weights for corresponding feature maps without gradient calculation. These gradient-based methods require only one forwar…”
Method
Research works printed but not linked: 12
- Arundel S.T., 2020, Transactions in GIS, 24, 556
- Herzfeld U.C., 2004, Atlas of antarctica: Topographic maps from geostatistical analysis of satellite radar altimeter data: with 169 figures
- Hesse R., 2021, Advances in Neural Information Processing Systems, 34
- Hsu C.Y., 2021, Remote Sensing, 13, 2116
- Janowicz K., 2022, AI Magazine, 43, 30
- Kim, B. et al., 2018. Interpretability beyond feature attribution: quantitative testing with concept activation vectors (TCAV). In: J. Dy and A. Krause, eds. International conference on machine learning. Stockholm, Sweden: PMLR, 2668–2677.
- Li W., 2022, GeoInformatica, 2022, 1
- Lin, M., Chen, Q., and Yan, S., 2014. Network in network. In: Y. Bengio and Y. LeCun, eds. 2nd International conference on learning representations, ICLR 2014, conference track proceedings, April 14–16, 2014. Banff, AB, Canada. https://sites.google.com/site/representationlearning2014/program-details/publication-model
- McEwen R.B., 1983, USGS digital cartographic data standards, 2
- Purcell A., 2018, Basic biology: An introduction
- Rieger, L. et al., 2020. Interpretations are useful: penalizing explanations to align neural networks with prior knowledge. In: L. Rieger, et al., eds. International conference on machine learning. PMLR, 8116–8126. http://proceedings.mlr.press/v119/rieger20a.html
- Springenberg, J.T. et al., 2015. Striving for simplicity: the all convolutional net. In: Y. Bengio and Y. LeCun, eds. 3rd International conference on learning representations, ICLR 2015, workshop track proceedings, May 7–9, 2015. San Diego, CA, USA. https://dblp.org/db/conf/iclr/iclr2015.htm
Cited by
Explainable spatially explicit geospatial artificial intelligence in urban analytics
Liu, P., Zhang, Y., & Biljecki, F. (2023) · Environment and Planning B Urban Analytics and City Science · doi:10.1177/23998083231204689 · Cited by 120
“Many of them fall in the domain of Transportation for traf fi c management and engineering-oriented purposes (Li, 2023; Wagner et al., 2022; Nascita et al., 2021a, 2021b), or there are other studies that have conducted XAI research with street view images (SVI) to study the street environment (Liu…”
BackgroundMethodAdvancing Arctic Sea Ice Remote Sensing with AI and Deep Learning: Opportunities and Challenges
Li, W., Hsu, C.-Y., & Tedesco, M. (2024) · Remote Sensing · doi:10.3390/rs16203764 · Cited by 83
Self-citationWhen Machine Learning Meets Geospatial Data: A Comprehensive GeoAI Review
Boutayeb, A., Lahsen-Cherif, I., & Khadimi, A. E. (2025) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2025.3568715 · Cited by 68
A five-year milestone: reflections on advances and limitations in GeoAI research
Hu, Y., Goodchild, M. F., Zhu, A., Yuan, M., Aydin, O., Bhaduri, B., Gao, S., Li, W., Lunga, D., & Newsam, S. D. (2024) · Annals of GIS · doi:10.1080/19475683.2024.2309866 · Cited by 61
“Some nice efforts have been made by researchers in this direction (Cheng et al. 2021; Z. Li 2022; Hsu and Li 2023; Xing and Sieber 2023).”
BackgroundSelf-citationMultimodal Colearning Meets Remote Sensing: Taxonomy, State of the Art, and Future Works
Kieu, N., Nguyen, K., Nazib, A., Fernando, T., Fookes, C., & Sridharan, S. (2024) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2024.3378348 · Cited by 51
“Among these techniques, feature attribution [172], [173] and visualization of model decisions [174], [175] are common within RS literature.”
Method
Metadata
- Authors
- Hsu, C.-Y., & Li, W.
- Affiliations
- Arizona State University (US)
- Corresponding author
- Li, Wenwen
- Venue
- International Journal of Geographical Information Science
- Publisher
- Informa UK Limited
- Published
- 2023-03-24
- Volume
- 37
- Issue
- 5
- Pages
- 963-987
- ISSN
- 0269-3798 (ISSN-L)
- 1365-8816
- 1362-3087
- OpenAlex
- W4360855232
- Cited by
- 75
- References
- 45
- Citations by year
- 2026: 23 · 2025: 27 · 2024: 21 · 2023: 4
- FWCI
- 7.07
- Type
- journal article
- Language
- en
- Open access
- green
- arxiv.org
- Licence
- cc-by-nc-sa
- Locations
- International Journal of Geographical Information Systems (publishedVersion)
- arXiv (Cornell University) (submittedVersion, cc-by-nc-sa)
- arXiv (Cornell University)
- Funders
- National Science Foundation (2120943, 2230034, 1853864)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Explainable Artificial Intelligence (XAI)
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Visual Attention and Saliency Detection
- Keywords
- saliency maps, model interpretability, perturbation-based explanation, gradient-based explanation, integrated gradients, class activation maps, object recognition
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
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Explainable GeoAI: can saliency maps help interpret artificial intelligence’s learning process? An empirical study on natural feature detection. https://app.cochat.ai/papers/10.1080/13658816.2023.2191256
- The paper
- Hsu, C.-Y., & Li, W. (2023). Explainable GeoAI: can saliency maps help interpret artificial intelligence’s learning process? An empirical study on natural feature detection. International Journal of Geographical Information Science. https://doi.org/10.1080/13658816.2023.2191256
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.