OSCR

A Comparative Analysis of Explainable AI (XAI) Techniques for Transparent and Reliable Image Classification.

Code ↔ Paper

2 matches between paragraphs of the paper and lines of its authors' code, computed by the harvester (lexical-v1). Click a colored paragraph or line to see its counterpart.

2 blocks of code here have a permanent identifier (oscr:b:…), listed with the matches below: citable for good, with a page of its own. Select a match and press Cite to copy it. What a block identifier is.

The 2 matches
  1. [1] § 3. Methodology › 3.3. Test Models: Role of CNN Architecture Complexity ↔ xai_c/models/definitions.py, lines 5 to 36 · score 0.73 · AvgPool2d, ReLU, kernel, module, stride, padding · block oscr:b:EFBWP56R8RN4Q-9 · Report a mismatch
  2. [2] § 3. Methodology › 3.2. Preprocessing of Dataset ↔ CNN_XAI_Analysis.py, lines 190 to 201 · score 0.60 · ToTensor, PIL image, resize, model · block oscr:b:RNC3RBXX4J729-L · Report a mismatch

Paper

Loaded from Europe PMC by your browser; OSCR does not store it: doi.org · Europe PMC

The paper is loaded when this pane is shown.

The authors' code

Python · 77 lines · 2.2 KB · Apache-2.0 · 1 match

  1. import torch
  2. import torch.nn as nn
  3. import torchvision.models as models
  4. # --- CNN Definition ---
  5. class CNN(nn.Module):
  6. def __init__(self):
  7. super(CNN, self).__init__()
  8. self.features = nn.Sequential(
  9. nn.Conv2d(3, 8, kernel_size=3, stride=1, padding=1),
  10. nn.ReLU(inplace=True),
  11. nn.AvgPool2d(kernel_size=2, stride=2),
  12. nn.Conv2d(8, 16, kernel_size=3, stride=1, padding=1),
  13. nn.ReLU(inplace=True),
  14. nn.AvgPool2d(kernel_size=2, stride=2),
  15. nn.Conv2d(16, 32, kernel_size=3, stride=1, padding=1),
  16. nn.ReLU(inplace=True),
  17. nn.AvgPool2d(kernel_size=2, stride=2),
  18. nn.Conv2d(32, 64, kernel_size=3, stride=1, padding=1),
  19. nn.ReLU(inplace=True),
  20. nn.AvgPool2d(kernel_size=2, stride=2)
  21. )
  22. self.classifier = nn.Sequential(
  23. nn.Linear(64 * 14 * 14, 64),
  24. nn.ReLU(inplace=True),
  25. nn.Linear(64, 2)
  26. )
  27. def forward(self, x):
  28. x = self.features(x)
  29. x = x.view(x.size(0), -1) # Flatten
  30. x = self.classifier(x)
  31. return x
  32. # --- Loaders ---
  33. def load_vgg16(weights_path, device):
  34. """
  35. Loads VGG16 with a binary classifier (Sigmoid output).
  36. """
  37. model = models.vgg16()
  38. # Modify classifier for binary output as requested
  39. model.classifier[6] = nn.Sequential(
  40. nn.Linear(4096, 1),
  41. nn.Sigmoid()
  42. )
  43. try:
  44. state_dict = torch.load(weights_path, map_location=device, weights_only=True)
  45. except TypeError:
  46. # Fallback for older pytorch versions
  47. state_dict = torch.load(weights_path, map_location=device)
  48. model.load_state_dict(state_dict)
  49. model.to(device)
  50. model.eval()
  51. return model
  52. def load_cnn(weights_path, device):
  53. """
  54. Loads Custom CNN with split state_dict (features/classifier).
  55. """
  56. model = CNN()
  57. # Load the dictionary of state dicts
  58. state_dict = torch.load(weights_path, map_location=device)
  59. # Load into respective parts
  60. model.features.load_state_dict(state_dict['features'])
  61. model.classifier.load_state_dict(state_dict['classifier'])
  62. model.to(device)
  63. model.eval()
  64. return model

definitions.py at commit a3145d6, under Apache-2.0 · at the source

Overview

  1. Department of Computer Science, Old Dominion University, Norfolk, VA 23529, USA; (S.C.); (S.M.D.)
  2. Environmental Laboratory, US Army Engineer Research and Development Center, Vicksburg, MS 39180, USA; (K.R.P.); (M.L.M.)
Institutions: Old Dominion University (United States); United States Army (United States); U.S. Army Engineer Research and Development Center (United States)
Journal: Entropy (Basel, Switzerland), volume 28, issue 5, article 562
Dates: received 12 February 2026; accepted 6 May 2026; published online 18 May 2026
Type: Research article · Language: English
License: CC BY
Identifiers: DOI 10.3390/e28050562 · PMID 42187977 · PMCID PMC13205289 · OpenAlex W7161676037
Open access: gold, a free copy (OpenAlex)
Status: code verified (how the code is found and verified)
Categories: methods / tools (subfield)
Methods: Connectivity, Machine learning
Keywords: XAI, image classification, LIME, Grad-CAM, PEEK, LRP, interpretability
Topic: Explainable Artificial Intelligence (XAI) (Artificial Intelligence, Computer Science), according to OpenAlex
Citations: not cited yet (Europe PMC); 39 references in the paper

Abstract

Evaluating the trustworthiness of black-box machine learning models remains a significant methodological challenge. Their lack of transparency and interpretability limits applicability, because stakeholders often seek transparency before trusting the results of black-box machine learning models. Explainable AI (XAI) methods provide for human-understandable justifications and informed decision-making of these black-box architectures. Therefore, it is imperative to select the proper XAI model tailored to specific tasks. In this research, we focus on examining four XAI techniques: PEEK, LRP, GRAD-CAM, and LIME to understand how they perform against each other for image classification tasks. We evaluate the performance, robustness, generalizability, noise stability, and computational efficiency of these methods using a globally recognized dataset. With 7390 images, the Oxford IIT pet dataset provides a comprehensive resource for training a custom Convolutional Neural Network (CNN) and VGG16, enabling a consistent evaluation of each XAI method. First, we analyze the saliency maps of the input images and observe the regions predicted by these XAI methods, and then leverage a noise analysis approach to evaluate their performance in terms of accuracy. We further explore the robustness, run-time, and “faithfulness” metrics of each XAI method. In general, we find that these methods can identify a set of input-data features that are critical for accurate classification but also intuitive, such as the outline, face, and eyes of subjects. However, our analysis reveals only marginal consensus among XAI methods in identifying those critical features. Grad-CAM demonstrates strong robustness and stability in VGG16, but the performance on the shallow CNN model remained inconsistent.

Reproduced under the paper's license (CC BY), from the paper cited above.

Repository

Its files are read in the Code ↔ Paper reader above, with 2 matches between paragraphs and lines of code.

AMLRL-ODU/CompareXAI

License: Apache-2.0
State: the link answers, verified on 28 September 2026
Evidence: files inventoried
Commit: a3145d624eac9ca8c9ed3c5001829707e8d3fe3e, 18 May 2026
Languages: Python (20)
Size: 23 files, 20 scripts
Software Heritage: not archived
Found in: “Data Availability Statement”
Holds: README, license file
Not found: CITATION.cff, environment file, tests, continuous integration, documentation
Tools: PyTorch (15 files), NumPy (11 files), Matplotlib (8 files), pandas (3 files), seaborn (3 files), OpenCV (2 files), Pillow (2 files), SciPy (2 files), scikit-learn (1 file)
Availability: 1 check, the latest on 28 September 2026: the link answers
  • 28 September 2026: the link answers
22 files

The paper's code and data availability statement is in the Data section.

Tracing map

Proposed by the machine: these links were found in the paper and verified at the source, without human review. The map will receive a Zenodo DOI once one of the paper's authors has validated it with their ORCID.

What the map holds:

  • 1 repository of the authors' code, each at its verified commit, with its license and how the link was found in the paper;
  • 20 scripts, each with its path and the digest of its content;
  • 2 matches between paragraphs of the paper and lines of the code (method lexical-v1);
  • neither the text of the paper nor the code itself.

Its JSON (tracing-map.json) is deposited on Zenodo with its DOI once the map is validated.

Data

No dataset and no data link were found in the paper.

Data Availability Statement

No new data were created or analyzed in this study. All the codes and data are available in the mentioned Github link: https://github.com/AMLRL-ODU/CompareXAI (accessed on 11 February 2026).

Reproduced under the paper's license (CC BY), from the paper cited above.

Versions

The history of this record: each version stored by the harvester or made by a correction of its authors or of the maintainers of its code, and what changed in its facts. The texts of the paper (its abstract, its availability statements) are not part of it; versions that changed only those are not listed.

Version 1, 28 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 5 authors, 7 keywords, 27 references.

Cite

This paper

Chakraborty, S., Dipto, S. M., Pilkiewicz, K. R., Mayo, M. L., & Rana, P. (2026). A Comparative Analysis of Explainable AI (XAI) Techniques for Transparent and Reliable Image Classification. Entropy (Basel, Switzerland), 28(5), 562. https://doi.org/10.3390/e28050562

BibTeX

@article{chakraborty2026comparative,
author = {Chakraborty, Sovon and Dipto, Shakib Mahmud and Pilkiewicz, Kevin R. and Mayo, Michael L. and Rana, Pratip},
title = {{A Comparative Analysis of Explainable AI (XAI) Techniques for Transparent and Reliable Image Classification}},
journal = {Entropy (Basel, Switzerland)},
year = {2026},
month = may,
volume = {28},
number = {5},
pages = {562},
publisher = {Multidisciplinary Digital Publishing Institute (MDPI)},
issn = {1099-4300},
doi = {10.3390/e28050562},
url = {https://doi.org/10.3390/e28050562},
pmid = {42187977},
pmcid = {PMC13205289}
}

RIS

TY - JOUR
AU - Chakraborty, Sovon
AU - Dipto, Shakib Mahmud
AU - Pilkiewicz, Kevin R.
AU - Mayo, Michael L.
AU - Rana, Pratip
TI - A Comparative Analysis of Explainable AI (XAI) Techniques for Transparent and Reliable Image Classification
T2 - Entropy (Basel, Switzerland)
J2 - Entropy (Basel)
PY - 2026
DA - 2026/05/18
VL - 28
IS - 5
SP - 562
SN - 1099-4300
PB - Multidisciplinary Digital Publishing Institute (MDPI)
DO - 10.3390/e28050562
UR - https://doi.org/10.3390/e28050562
LA - en
ER -

CSL-JSON

{
"id": "10.3390/e28050562",
"type": "article-journal",
"title": "A Comparative Analysis of Explainable AI (XAI) Techniques for Transparent and Reliable Image Classification",
"container-title": "Entropy (Basel, Switzerland)",
"author": [
{
"family": "Chakraborty",
"given": "Sovon"
},
{
"family": "Dipto",
"given": "Shakib Mahmud"
},
{
"family": "Pilkiewicz",
"given": "Kevin R."
},
{
"family": "Mayo",
"given": "Michael L."
},
{
"family": "Rana",
"given": "Pratip"
}
],
"container-title-short": "Entropy (Basel)",
"volume": "28",
"issue": "5",
"page": "562",
"DOI": "10.3390/e28050562",
"PMID": "42187977",
"PMCID": "PMC13205289",
"ISSN": "1099-4300",
"publisher": "Multidisciplinary Digital Publishing Institute (MDPI)",
"URL": "https://doi.org/10.3390/e28050562",
"language": "en",
"issued": {
"date-parts": [
[
2026,
5,
18
]
]
}
}

The tracing map gets a citation of its own once an author has validated it and it has a DOI.

Similar papers

The papers with a page that share the most with this one: the tools found in their code, their categories, datasets, cited references and authors, the rarest counting most.

[1] doi:10.1016/j.neuroimage.2023.120109 [code]
Benchmarking explanation methods for mental state decoding with deep learning models.
Journal: NeuroImage
In common: Pillow, PyTorch, seaborn, 5 other tools, methods / tools, 1 reference
[2] doi:10.3389/frai.2022.903875 [code]
Benchmarking Perturbation-Based Saliency Maps for Explaining Atari Agents.
Journal: Frontiers in artificial intelligence
In common: OpenCV, Pillow, seaborn, 5 other tools, 1 reference
[3] doi:10.1007/s10994-022-06167-y [code]
Scrutinizing XAI using linear ground-truth data with suppressor variables.
Journal: Machine learning
In common: Pillow, seaborn, scikit-learn, 4 other tools, methods / tools, 1 reference
[4] doi:10.1038/s41598-025-09538-2 [code]
A comprehensive analysis of perturbation methods in explainable AI feature attribution validation for neural time series classifiers.
Journal: Scientific reports
In common: PyTorch, seaborn, scikit-learn, 4 other tools, methods / tools, 1 reference
[5] doi:10.1038/s41598-026-47814-x [code]
Predicting post-stroke functional outcome using explainable machine learning and integrated data.
Journal: Scientific reports
In common: PyTorch, scikit-learn, pandas, 2 other tools, 2 references
[6] doi:10.1109/jtehm.2024.3430035 [code]
XAI-Based Assessment of the AMURA Model for Detecting Amyloid-β and Tau Microstructural Signatures in Alzheimer's Disease.
Journal: IEEE journal of translational engineering in health and medicine
In common: Pillow, PyTorch, scikit-learn, 4 other tools, 1 reference
[7] doi:10.3390/e28040453 [code]
Polynomial Perceptrons for Compact, Robust, and Interpretable Machine Learning Models.
Journal: Entropy (Basel, Switzerland)
In common: PyTorch, seaborn, scikit-learn, 4 other tools, 1 reference
[8] doi:10.3389/frobt.2024.1324404 [code]
Neural dynamics of robust legged robots.
Journal: Frontiers in robotics and AI
In common: OpenCV, Pillow, PyTorch, 3 other tools, 1 reference
[9] doi:10.1038/s41746-024-01123-7 [code]
Constructing personalized characterizations of structural brain aberrations in patients with dementia using explainable artificial intelligence.
Journal: NPJ digital medicine
In common: Pillow, scikit-learn, pandas, 3 other tools, 1 reference
[10] doi:10.3389/fpsyt.2026.1803720 [code]
NeuroCon-AutismNet: a privacy-preserving multimodal framework toward autism screening via diffusion-regularized EEG biomarkers and empathy-aware multilingual dialogue.
Journal: Frontiers in psychiatry
In common: PyTorch, scikit-learn, pandas, 3 other tools, 1 reference

Contribute

The authors of this paper can claim it, correct its record and validate its tracing map, and the maintainers of its code (its owner, or a public member of its organization) correct what it says of their repository; anyone signed in can ask for its removal. Every request goes to OSCR's own machine, which answers it; your account page follows them.

Sign in with ORCID to claim this paper as one of its authors, correct its record or validate its tracing map: when the paper's metadata lists your ORCID iD, you are recognized at once. Maintainers of its code: sign in with GitHub, then claim the repository on your account page.

Add the badge to the code's README

OSCR: the badge of this paper

If you are an author or a maintainer of this paper's code, you can add this badge. It says what OSCR knows of the paper (its code, its validated tracing map) and links the code to this page. Paste one of these snippets under the README's title. Keep the two marks around it, so that the badge can be replaced in place and is never added twice. OSCR changes nothing in your repository.

Markdown
<!-- oscr-badge:start -->
[![OSCR](https://openscicode.org/badge/doi/10.3390/e28050562.svg)](https://openscicode.org/badge/doi/10.3390/e28050562)
<!-- oscr-badge:end -->
reStructuredText
.. oscr-badge:start

.. image:: https://openscicode.org/badge/doi/10.3390/e28050562.svg
   :target: https://openscicode.org/badge/doi/10.3390/e28050562
   :alt: OSCR

.. oscr-badge:end
HTML
<!-- oscr-badge:start -->
<a href="https://openscicode.org/badge/doi/10.3390/e28050562"><img src="https://openscicode.org/badge/doi/10.3390/e28050562.svg" alt="OSCR"></a>
<!-- oscr-badge:end -->

On GitHub, open the README, choose Edit (the pencil), paste the snippet under the title, then "Commit changes".

If you choose to, you can have the badge kept up to date for you, with oscr badge --write from the command line or with the GitHub Action in your repository's workflows. Every other form and option (a card, AsciiDoc, LaTeX, shields.io) is on the badge page.

Request its removal

To ask OSCR to remove this record, the copies of its authors' scripts or its tracing map, use the removal request page: signed in, you say who you are, what to remove and why, then review and confirm the request. Published rules decide every request (how).

Discussion, reproductions, activity

Discussion: code errors and mismatches between the code and the paper about this paper and its code, from readers and its authors, beside the paper and its tracing map.

Report a mismatch between the code and the paper · Report a code error (signed in)

Reproductions: reports from readers who ran the authors' code and did not get the paper's result, what they ran, with which environment, commit and data.

Report a reproduction failure (signed in)

Versions of its code: the releases of its code tied to a version of this paper (the preprint, the accepted manuscript, the version of record…), each with the tracing map versioned with it.

Activity: what happens around this paper: new versions of its record, its map's validation, discussions and reproductions. It opens with sign-in.