Accelerating Post-Tornado Disaster Assessment Using Advanced Deep Learning Models

Robinson Umeike; Shane Crawford; Thang Dao

REVIEW 4 major objections 5 minor 14 references

Accelerating Post-Tornado Disaster Assessment Using Advanced Deep Learning Models

T0 review · 4 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A dual deep-learning pipeline automates post-tornado damage triage, with ResNet50 at 90.28% accuracy and YOLOv11 detecting buildings in 3 ms per frame.

desk verdict Applied tornado-damage classification with standard models; the numbers are plausible but the evaluation needs a building-aware split before they can be trusted. read the letter →

arxiv 2412.18147 v1 pith:L2LJE4AT submitted 2024-12-24 cs.CV

Robinson Umeike , Thang Dao , Shane Crawford This is my paper · ORCID

classification cs.CV

keywords damageassessmentcomputervisionresiliencepost-tornadodisasterdeeplearningYOLOv11ResNet50transfer

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper claims that a two-model computer vision system can automate post-tornado building damage assessment, replacing weeks of manual inspection with near-real-time analysis. On image sets from the 2013 Moore, Oklahoma and 2021 Mayfield, Kentucky tornadoes, the authors report that a ResNet50 classifier reaches 90.28% accuracy in assigning one of five damage states, while a YOLOv11 detector locates buildings at 3 ms per frame with 60.83% accuracy. If correct, this would give disaster responders a fast, consistent, and scalable triage tool for a disaster type that prior AI work has largely skipped. The figures are claimed for the specific datasets and splits described in the paper, not for all tornado imagery.

What carries the argument

The machinery is a two-part architecture. For damage classification, a ResNet50 backbone pretrained on ImageNet is stripped of its top layers and fitted with a global average pooling 2D layer, a dense layer of 1024 units with ReLU activation, and a softmax output; categorical cross-entropy loss drives the training. For building detection, YOLOv11 runs the entire image through a single forward pass to output bounding boxes, class labels, and confidence scores. Supporting this are a data pipeline that auto-orients and resizes frames, augmentation with rotations, shifts, shear, zoom, and flips, and two labeled datasets: hand-held camera images of the 2013 Moore tornado and 360° vehicle-video frames of the 2021 Mayfield tornado. The ResNet50 transfer-learning setup carries the classification accuracy, while the YOLOv11 single-pass design carries the real-time detection speed.

What would settle it

Re-run both evaluations on a test set constructed by grouping all frames from the same building or the same 360° video sequence into one side of the split, so no near-duplicate frames cross the train/test boundary; if classification accuracy falls materially below 90.28% or detection accuracy below 60.83%, the reported numbers are inflated by temporal correlation.

Watch

Extended reading notes

Core claim

The authors claim that a dual-model deep learning system—ResNet50 for fine-grained damage classification and YOLOv11 for building detection and localization—can perform post-tornado damage assessment at a level suitable for practical response workflows. On 2,635 images from the 2013 Moore tornado, the ResNet50 model distinguishes undamaged buildings, roof damage, wall collapse, wall-and-roof damage, and non-buildings with 90.28% accuracy at 1,529 ms per image. On 1,776 frames extracted from 360° vehicle video of the 2021 Mayfield tornado, YOLOv11 detects buildings with 60.83% accuracy, a mean average precision of 0.45 at 50% IoU, and a 3.0 ms inference time. The authors present this as a scalable, objective alternative to manual post-disaster assessment, and as the first tornado-specific application in a literature dominated by earthquake, flood, and hurricane damage.

Load-bearing premise

The reported accuracies depend on the test images being genuinely new to the models, meaning no near-duplicate frames from the 360° video and no second images of the same building appear in both the training and test sets.

Editorial extensions

If this is right

Post-tornado damage assessment could drop from weeks of manual inspection to near-real-time analysis of drive-by video and still images.
The same dual-model pipeline could be retrained on damage datasets from other hazard types, since the underlying architectures are generic.
YOLOv11's 3 ms per frame makes it feasible to run building detection live during vehicle-mounted surveys, flagging damaged structures as they are filmed.
Automated damage-state labeling would give authorities a consistent, repeatable damage assessment that is not subject to human evaluator bias.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

If the reported accuracy holds on a split that is de-duplicated at the building or video-sequence level, the approach would likely transfer to other windstorm events; a direct test would be training on one tornado and evaluating on a later, unseen tornado.
The large gap between the 90.28% classification accuracy and the 60.83% detection accuracy suggests that building localization, not damage-state classification, is the current bottleneck, so improving the detector may yield larger system gains than further classifier tuning.
The 360° video source almost certainly produces temporally correlated frames, so a sequence-aware train/test split could materially change the reported numbers; reporting such a split would be the most direct stress test of the 90.28% figure.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

First thing to know: this is an applied-transfer paper, not a methods paper. The new content is two tornado damage datasets (Moore 2013, Mayfield 2021) and a pair of off-the-shelf models run on them. The architecture, training, and evaluation are all standard. That is fine as a contribution, but the headline accuracy needs to be read with caution.

The paper does several things right. It describes data curation and preprocessing in enough detail to reproduce the pipeline. The two tasks—building detection and damage classification—are genuinely different and each uses a sensible model. The authors are open about class imbalance and future work. The citation pattern is normal; they cite the prior data-collection papers and the YOLO/ResNet sources.

The soft spots are in the evaluation. The 90.28% accuracy comes from a random 80:10:10 split of images. The Moore data was gathered with handheld cameras, so the same building almost certainly appears in multiple photos. If those photos land in both train and test, the model can memorize building appearance and overstate accuracy. The paper never mentions grouping by building, sequence, or near-duplicate removal. That is the main threat to the central claim. Also missing: per-class precision/recall (the text says they're in Figure 4 but doesn't give numbers), error bars or repeated runs, and any baseline comparison (even a simple ResNet without augmentation would help). The 1529ms per-image inference time is oddly high for ResNet50 on an RTX 4090—typical is tens of milliseconds—and is not explained. The YOLOv11 result (60.83% accuracy, mAP@0.5≈0.45) is modest; calling it 'rapid detection' is fair, but it does not support near-expert triage.

These are not fatal flaws. The paper is a legitimate application study and the domain gap for tornado damage is real. The right fix is a building-aware split and more careful reporting. That is a major revision, not a desk reject. I would send it to a knowledgeable reviewer and ask for those changes, plus code/data release so others can build on the datasets.

Referee Report

4 major / 5 minor

Summary. The paper proposes a dual-model deep learning pipeline for post-tornado building damage assessment. A ResNet50 classifier with transfer learning is trained on 2,635 images from the 2013 Moore tornado to categorize damage into five classes, reporting 90.28% accuracy and a 1529 ms per-image inference time; a YOLOv11 detector is trained on 1,776 frames extracted from 360° vehicle video of the 2021 Mayfield tornado, reporting 60.83% accuracy and 3 ms inference time. The authors argue that this automated approach can accelerate manual damage assessment and improve disaster response.

Significance. If the reported results are valid, the paper offers a practical application of standard deep learning to tornado damage assessment, a comparatively underexplored disaster type in the computer-vision literature. The use of two complementary tasks—object detection of buildings and fine-grained damage classification—on real post-disaster data is a strength, as is the reporting of inference times on a modern GPU. However, the evidence is weakened by evaluation-methodology gaps: no building-disjoint or sequence-aware split, no per-class metrics for a heavily imbalanced classification problem, an unexplained and implausible inference-time figure for ResNet50, and an undefined 'accuracy' metric for the detector. These gaps are load-bearing because the headline claims (near-expert accuracy and near-real-time speed) depend on them. The central idea is plausible, but the paper needs stronger validation before the claims can be accepted.

major comments (4)

[§IV, Table 1] The 90.28% accuracy claim for the ResNet50 classifier is not supported as a generalization estimate because the train/validation/test split is performed at the image level, not grouped by building or de-duplicated. The Moore dataset (Section II.A) was collected with handheld cameras, which typically produce multiple images of the same building from different angles. If the same building appears in both training and test partitions, the model can memorize building-specific textures and inflate accuracy. The manuscript describes no building-level or sequence-aware splitting and no analysis of near-duplicate frames. Please evaluate with a building-disjoint split (e.g., grouping images by building or by geographic proximity) and report the resulting accuracy, or provide evidence that no building appears in more than one partition.
[§IV, Table 1, Fig. 2] For the heavily imbalanced Moore classification task, overall accuracy alone is insufficient. The bottom panel of Figure 2 indicates that the 'undamaged building' class dominates, and the paper acknowledges class imbalance in Section V. A model can achieve high overall accuracy by predicting the majority class for most inputs. The manuscript claims per-class metrics are shown in Figure 4 but does not report any numeric per-class precision, recall, or macro-F1 in the text. Please include a confusion matrix with class-wise precision/recall and a macro-averaged F1 score to substantiate the claim of near-expert performance.
[Abstract, Table 1] The reported inference time of 1529 ms per image for ResNet50 on an NVIDIA GeForce RTX 4090 is conspicuously high; a standard ResNet50 forward pass on this GPU is expected to take only a few milliseconds. If this number includes image loading, preprocessing, and Python overhead, that must be stated explicitly and separated from the model's forward-pass time. As written, the 'near real time analysis' claim in Section V is not quantitatively supported. Please clarify the measurement protocol or correct the value.
[§IV, Table 1] The 'accuracy' metric for the YOLOv11 detection model is ambiguous. For an object detector, accuracy is not a standard metric; the paper reports mAP@0.5 ≈ 0.45 and mAP@0.5:0.95 ≈ 0.27, then states an overall accuracy of 60.83% without defining what counts as a correct prediction (per-image, per-bounding-box, or at what IoU threshold). Please define the accuracy computation and reconcile it with the mAP values, or replace it with standard detection metrics such as mAP and precision/recall.

minor comments (5)

[§II.B.1, Table 1] There are several typos: 'a ny newly created pixels' should be 'any newly created pixels', 'Data Proprocessing' should be 'Data Preprocessing', and 'Training Paramaters' should be 'Training Parameters'.
[§III.B] The cross-entropy equation has rendering artifacts in the submitted PDF (garbled subscripts and formatting). Please ensure the equation is typeset correctly.
[§IV, Figures 3 and 4] The text states that all evaluation metrics are summarized in Figures 3 and 4, but it does not refer to specific panels when discussing each metric. Please reference the relevant panels in the body and ensure the figures are legible in the final version.
[General] No data or code availability statement is provided. To support reproducibility, the authors should state whether the datasets or the trained models will be released, or at least describe the data-collection protocol in sufficient detail to allow re-collection.
[§I, References] The YOLOv11 model is attributed to Redmon et al. (the original YOLO paper), but YOLOv11 is a different implementation with its own publication. Please cite the correct YOLOv11 source or, at minimum, a repository or technical report that defines the architecture used.

Circularity Check

0 steps flagged · score 2.0 of 10

No circularity: 90.28% accuracy is an empirical held-out test result; the sole self-citation is data provenance, not a load-bearing premise.

full rationale

This is an empirical supervised-learning paper rather than a derivation chain, so the main circularity patterns do not apply. The headline ResNet50 accuracy (90.28%) is reported as the accuracy on the 10% test split after an 80:10:10 split (Table 1), with the model trained on the 80% training portion; no test-set labels enter the training objective, and no fitted parameter is renamed as a prediction. The YOLOv11 result (60.83%) is similarly evaluated on its own 73:18:9 split. The only author self-reference is the provenance of the Mayfield dataset: 'we extracted images of damaged buildings from vehicle-mounted, 360° video data collected in Mayfield, Kentucky following the data collection and processing method in [2, 9].' Reference [9] is a prior paper by coauthor Crawford, but it supplies a data-collection procedure, not a theorem or fitted value, and does not by itself force either accuracy number; each result is empirically measured on held-out data and is falsifiable. No uniqueness theorem, imported ansatz, or renamed known result is invoked. The reader's and skeptic's concerns about random splitting, possible near-duplicate frames, and class imbalance are validity and generalization risks, not circularity: they question whether the test accuracy would transfer to new buildings, not whether the reported number reduces to its input by construction. Given the single minor, non-load-bearing self-citation, the appropriate score is 2.

Assumptions & free parameters 1 free parameters · 4 assumptions · 0 invented entities

The central empirical claim rests on the quality of manual labels, the independence of the data split from video-frame redundancy, the validity of ImageNet transfer, and the i.i.d. assumption. The training hyperparameters are chosen by hand and are listed as a free parameter because the reported accuracy is conditional on them. No new entities are introduced.

free parameters (1)

Training hyperparameters (learning rate, batch size, epochs, augmentation ranges)
These values are selected by the authors rather than tuned on the test set, but the reported accuracy depends on them and no ablation or sensitivity analysis is provided (Section II.B and Table I).

assumptions (4)

domain assumption Human-annotated damage labels are correct and consistent across both datasets.
All training and test labels come from manual tagging of damage states; errors in these labels propagate directly into the reported accuracies (Section II.A).
domain assumption Train/validation/test splits are independent, with no near-duplicate frames of the same building across splits.
The 2021 Mayfield data come from vehicle-mounted 360-degree video; without de-duplication, temporally adjacent frames of the same building could appear in both training and test sets, inflating accuracy (Section II.A and Table I).
domain assumption ImageNet-pretrained ResNet50 features transfer to tornado damage imagery.
The model uses ImageNet weights and a new classifier head; no analysis is given for whether this pretraining is appropriate for the damage domain (Section III.B).
domain assumption The evaluation distribution matches the deployment distribution for new tornadoes.
The test set is a random split from the same two events, not a novel event; claims of generalized scalability assume future tornado imagery is similar (Section II.A and Section V).

how reviews work

0 comments

Cite this review

Pith. "Pith review of Accelerating Post-Tornado Disaster Assessment Using Advanced Deep Learning Models." pith.science (2026). https://pith.science/paper/L2LJE4AT

@misc{pith2026241218147,
  author       = {Pith},
  title        = {Pith review of: Accelerating Post-Tornado Disaster Assessment Using Advanced Deep Learning Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/L2LJE4AT}},
  note         = {Machine review of arXiv:2412.18147}
}

read the original abstract

Post-disaster assessments of buildings and infrastructure are crucial for both immediate recovery efforts and long-term resilience planning. This research introduces an innovative approach to automating post-disaster assessments through advanced deep learning models. Our proposed system employs state-of-the-art computer vision techniques (YOLOv11 and ResNet50) to rapidly analyze images and videos from disaster sites, extracting critical information about building characteristics, including damage level of structural components and the extent of damage. Our experimental results show promising performance, with ResNet50 achieving 90.28% accuracy and an inference time of 1529ms per image on multiclass damage classification. This study contributes to the field of disaster management by offering a scalable, efficient, and objective tool for post-disaster analysis, potentially capable of transforming how communities and authorities respond to and learn from catastrophic events.

Figures

Figures reproduced from arXiv: 2412.18147 by the authors.

**Figure 1.** Modified ResNet50 Architecture for Transfer Learning [4]. Input Layer (224x224x3) 7x7 Conv, 64, Stride 2 3x3 max pool, stride 2 1x1 Conv, 64 3x3 Conv, 64 1x1 Conv, 256 1x1 Conv, 64 3x3 Conv, 64 1x1 Conv, 256 1x1 Conv, 64 3x3 Conv, 64 1x1 Conv, 256 1x1 Conv, 64 3x3 Conv, 64 1x1 Conv, 256 1x1 Conv, 128 3x3 Conv, 128 1x1 Conv, 512 1x1 Conv, 64 3x3 Conv, 64 1x1 Conv, 256 1x1 Conv, 64 3x3 Conv, 64 1x1 Conv, 256 1x1 Conv,… view at source ↗

**Figure 2.** Damage class distribution: Top: 2021 Midwest Tornado (Mayfield, KY), Bottom: 2013 Moore Tornado. II. METHODS A. Data Curation and Preprocessing The foundation of our research lies in a diverse dataset compiled from two significant tornado events (The 2021 Midwest Tornado and the 2013 Moore Tornado). For the former, we extracted images of damaged buildings from vehicle-mounted, 360° video data collected in Mayfield, … view at source ↗

**Figure 4.** Evaluation Metrics (ResNet50): Learning Curves (Model Accuracy and Loss), Precision-Recall Curve, ROC Curve, and the Confusion Matrix V. CONCLUSION This research advances post-disaster building assessment through innovative deep learning solutions. Our dual-model approach demonstrates compelling results: ResNet50 achieved 90.28% accuracy in damage classification (1529ms/image), while YOLOv11 enabled rapid detection … view at source ↗

Figures from the paper (1 more)

**Figure 3.** Figure 3: Evaluation Metrics (YOLOv11): Training Losses, Precision, Recall, Mean Average Precision, and Sample Predictions with Confidence Scores [PITH_FULL_IMAGE:figures/full_fig_p003_3.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

14 extracted references · 9 canonical work pages

[1]

Damage Classification: The i mages were organized into directories corresponding to their respective damage classification labels. To ensure consistent input dimensions for our deep learning model, all images were resized to 224x224 pixels with three color channels (RGB), resulting in an input shape of (224, 224, 3)

work page
[2]

Damage Detection: First, we applied an auto - orientation procedure to correct EXIF metadata orientation inconsistencies in the source images. This technical correction ensures that all images are read with their pixels properly aligned with their intended viewing orientation , preventing inadvertent misalignment between image data and model training. Wit...

work page
[3]

These techniques were applied dynamically during the training process through a custom data generator

Damage Classification: Data augmentation methods were employed to boost the robustness of our model and mitigate overfitting. These techniques were applied dynamically during the training process through a custom data generator. The generator applied rotations of up to 40 degrees, width and height shifts up to 20% of the image dimensions, as well as shear...

work page
[4]

This approach enhances the model's ability to accurately detect damage across different contexts

Damage Detection: Random rotations within a -10% to +10% range were applied to the dataset to improve the model's generalization to minor distortions, particularly those arising from extracting images from 360° video footage, as well as variations encountered in real -world scenarios. This approach enhances the model's ability to accurately detect damage ...

work page 2000
[5]

Lisa” Wang et al., “Application of Multidisciplinary Community Resilience Modeling to Reduce Disaster Risk: Building Back Better,

W. “Lisa” Wang et al., “Application of Multidisciplinary Community Resilience Modeling to Reduce Disaster Risk: Building Back Better,” Journal of Performance of Constructed Facilities, vol. 38, no. 3. American Society of Civil Engineers (ASCE), Jun. 2024. doi: 10.1061/jpcfev.cfeng-4650

work page doi:10.1061/jpcfev.cfeng-4650 2024
[6]

Interdisciplinary data collection for empirical community-level recovery modelling,

B. Johnston et al., “Interdisciplinary data collection for empirical community-level recovery modelling,” IABSE Reports, vol. 120. International Association for Bridge and Structural Engineering (IABSE), pp. 1260–1267, 2024. doi: 10.2749/manchester.2024.1260

work page arXiv 2024
[7]

Turning Disaster into Knowledge in Geotechnical Earthquake Engineering,

J. D. Bray, J. D. Frost, E. M. Rathje, and F. E. Garcia, “Turning Disaster into Knowledge in Geotechnical Earthquake Engineering,” Geotechnical Earthquake Engineering and Soil Dynamics V. American Society of Civil Engineers, pp. 186 –200, Jun. 07, 2018. doi: 10.1061/9780784481462.018

work page doi:10.1061/9780784481462.018 2018
[8]

An uncertainty-aware framework for reliable disaster damage assessment via crowdsourcing,

A. B. Khajwal and A. Noshadravan, “An uncertainty-aware framework for reliable disaster damage assessment via crowdsourcing,” International Journal of Disaster Risk Reduction, vol. 55. Elsevier BV, p. 102110, Mar. 2021. doi: 10.1016/j.ijdrr.2021.102110

arXiv 2021

Show all 14 references

[9]

A systematic review of trustworthy artificial intelligence applications in natural disasters,

A. S. Albahri et al., “A systematic review of trustworthy artificial intelligence applications in natural disasters,” Computers and Electrical Engineering, vol. 118. Elsevier BV, p. 109409, Sep. 2024. doi: 10.1016/j.compeleceng.2024.109409

2024
[10]

Building structural analysis based Internet of Things network assisted earthquake detection,

R. Duggal et al., “Building structural analysis based Internet of Things network assisted earthquake detection,” Internet of Things, vol. 19. Elsevier BV, p. 100561, Aug. 2022. doi: 10.1016/j.iot.2022.100561

2022
[11]

You Only Look Once: Unified, Real-Time Object Detection,

J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, “You Only Look Once: Unified, Real-Time Object Detection,” 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, pp. 779– 788, Jun. 2016. doi: 10.1109/cvpr.2016.91

2016 doi
[12]

Deep Residual Learning for Image Recognition,

K. He, X. Zhang, S. Ren, and J. Sun, “Deep Residual Learning for Image Recognition,” 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Jun. 2016. doi: 10.1109/cvpr.2016.90

2016 doi
[13]

Rapid Disaster Data Dissemination and Vulnerability Assessment through Synthesis of a Web‐Based Extreme Event Viewer and Deep Learning,

P. S. Crawford,“Rapid Disaster Data Dissemination and Vulnerability Assessment through Synthesis of a Web‐Based Extreme Event Viewer and Deep Learning,” Advances in Civil Engineering, vol. 2018, no. 1. Wiley, Jan. 2018. doi: 10.1155/2018/7258156

2018 doi
[14]

Tornado damage assessment in the aftermath of the May 20th 2013 Moore Oklahoma tornado,

Graettinger, Andrew, et al. "Tornado damage assessment in the aftermath of the May 20th 2013 Moore Oklahoma tornado," 2014

2013

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.

[1] [1]

Damage Classification: The i mages were organized into directories corresponding to their respective damage classification labels. To ensure consistent input dimensions for our deep learning model, all images were resized to 224x224 pixels with three color channels (RGB), resulting in an input shape of (224, 224, 3)

work page

[2] [2]

Damage Detection: First, we applied an auto - orientation procedure to correct EXIF metadata orientation inconsistencies in the source images. This technical correction ensures that all images are read with their pixels properly aligned with their intended viewing orientation , preventing inadvertent misalignment between image data and model training. Wit...

work page

[3] [3]

These techniques were applied dynamically during the training process through a custom data generator

Damage Classification: Data augmentation methods were employed to boost the robustness of our model and mitigate overfitting. These techniques were applied dynamically during the training process through a custom data generator. The generator applied rotations of up to 40 degrees, width and height shifts up to 20% of the image dimensions, as well as shear...

work page

[4] [4]

This approach enhances the model's ability to accurately detect damage across different contexts

Damage Detection: Random rotations within a -10% to +10% range were applied to the dataset to improve the model's generalization to minor distortions, particularly those arising from extracting images from 360° video footage, as well as variations encountered in real -world scenarios. This approach enhances the model's ability to accurately detect damage ...

work page 2000

[5] [5]

Lisa” Wang et al., “Application of Multidisciplinary Community Resilience Modeling to Reduce Disaster Risk: Building Back Better,

W. “Lisa” Wang et al., “Application of Multidisciplinary Community Resilience Modeling to Reduce Disaster Risk: Building Back Better,” Journal of Performance of Constructed Facilities, vol. 38, no. 3. American Society of Civil Engineers (ASCE), Jun. 2024. doi: 10.1061/jpcfev.cfeng-4650

work page doi:10.1061/jpcfev.cfeng-4650 2024

[6] [6]

Interdisciplinary data collection for empirical community-level recovery modelling,

B. Johnston et al., “Interdisciplinary data collection for empirical community-level recovery modelling,” IABSE Reports, vol. 120. International Association for Bridge and Structural Engineering (IABSE), pp. 1260–1267, 2024. doi: 10.2749/manchester.2024.1260

work page arXiv 2024

[7] [7]

Turning Disaster into Knowledge in Geotechnical Earthquake Engineering,

J. D. Bray, J. D. Frost, E. M. Rathje, and F. E. Garcia, “Turning Disaster into Knowledge in Geotechnical Earthquake Engineering,” Geotechnical Earthquake Engineering and Soil Dynamics V. American Society of Civil Engineers, pp. 186 –200, Jun. 07, 2018. doi: 10.1061/9780784481462.018

work page doi:10.1061/9780784481462.018 2018

[8] [8]

An uncertainty-aware framework for reliable disaster damage assessment via crowdsourcing,

A. B. Khajwal and A. Noshadravan, “An uncertainty-aware framework for reliable disaster damage assessment via crowdsourcing,” International Journal of Disaster Risk Reduction, vol. 55. Elsevier BV, p. 102110, Mar. 2021. doi: 10.1016/j.ijdrr.2021.102110

arXiv 2021