Pith. sign in

REVIEW 4 major objections 4 minor 35 references

Assessing the Benefits of Combining Advanced Deep Learning Techniques for Post-Disaster Building Damage Assessment from UAV Imagery

T0 review · 4 major / 4 minor · reviewed 2026-08-04 · deepseek-v4-flash

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

Pith's one-line read A two-stage pipeline—computer-vision detection feeding bounding boxes to a vision-language model—can count post-disaster building damage more accurately than either approach alone, even with as few as 100 labeled training images for the det

desk verdict Useful empirical study of a known hybrid pattern in a new domain, but the headline claim of consistent outperformance is contradicted by the paper's own Table 4 and needs tempering. read the letter →

arxiv 2608.01906 v1 pith:5IFPA4KC submitted 2026-08-03 cs.CV

Huy Quang Ung , Guillaume Habault , Roberto Legaspi , Hao Niu , Lian Cao , Masato Taya This is my paper
classification cs.CV
keywords post-disasterdamageassessmentUAVimagerybuildingcountinglargevision-languagemodelsGroundingDINOhybridCV-LVLMframeworkRescueNetFlood
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

Post-disaster damage assessment from drone imagery needs both precise localization of individual buildings and contextual judgment about what counts as damaged. The paper's central claim is that these should be split: let a computer-vision detector (Grounding DINO) draw the building boxes, then let a large vision-language model look at those boxes and reason about each building's damage level. On RescueNet and FloodNet, the best hybrid configuration raises counting accuracy by up to 2.1 R² points over isolated baselines, while the detector needs only 100 labeled images to fine-tune. The paper also reports the gain is not universal: on FloodNet, adding detector boxes sometimes hurts open-source vision-language models, and it identifies dataset quirks—sequence leakage and annotation inconsistencies—that qualify the results. If correct, the framework offers a practical low-data route to automated damage counting for emergency response.

What carries the argument

The central mechanism is a two-stage decoupling: Stage 1 uses the open-vocabulary detector Grounding DINO to localize buildings and emit bounding boxes with confidence scores; Stage 2 passes the image with overlaid boxes plus a class-specific counting prompt to an LVLM, which classifies each boxed region and returns a count. The identity doing the work is the bounding box as a spatial proposal: it supplies precise where-information while preserving the image content inside the box, so the LVLM can reason over damage cues without needing to detect. The paper's ablation shows rectangular boxes beat polygonal boxes and pixel masks, because masks obscure the very surfaces that reveal damage.

What would settle it

A decisive check: take a UAV disaster dataset with per-building damage labels and compare (a) the LVLM alone, (b) the LVLM+G-DINO hybrid, and (c) the LVLM with ground-truth boxes. If (c) does not beat (a) by a meaningful margin, or if (b) is no better than (a) whenever the detector's AP75 is high, then the central premise—that detection quality is the limiting factor—is wrong. The paper's own Table 4 already shows such a case: InternVL on FloodNet reaches R²=0.758 alone, but drops to R²=0.399 when combined with G-DINO fine-tuned on 100 samples.

Watch

Extended reading notes

Core claim

On its own terms, the paper establishes that decoupling detection from damage assessment yields a counting system that beats both a pure detector (G-DINO, fine-tuned to classify as well as detect) and pure LVLMs (QwenVL, InternVL, GPT, Gemini) on two real UAV benchmarks. The authors' strongest reported result is Gemini 3 Pro + G-DINO with 100-sample fine-tuning on RescueNet, surpassing the Gemini-alone baseline by 1.6 R² and G-DINO* by 2.1 R². They argue this works because bounding boxes give the reasoning model a precise spatial anchor without forcing the LVLM to do low-level localization it is bad at; their zero-shot detection table shows LVLM AP75 near zero while G-DINO reaches roughly 0.

Load-bearing premise

The load-bearing premise is that supplying a vision-language model with a detector's bounding boxes improves damage counting compared with the LVLM working alone; the paper's FloodNet results show this premise fails for some open-source models, so the framework's success depends on the specific model–dataset pairing.

Editorial extensions

If this is right

  • With only 100 labeled images for the detector, the hybrid already approaches full-data performance on both datasets, suggesting low-annotation deployments are feasible.
  • Improving detector box quality (higher AP75) generally improves counting, so investing in better detection transfers directly to more accurate damage counts.
  • Rectangular bounding boxes are better inputs for LVLM reasoning than segmentation masks, so practitioners should prefer simple boxes over precise masks for this kind of assessment.
  • The framework's benefit is not uniform: on FloodNet open-source LVLMs can degrade when boxes are added, so a system designer should benchmark both the LVLM alone and the hybrid before committing.
  • The paper's finding that images from the same flight sequence leak across splits and that FloodNet's annotations deviate from its own definition means current benchmark numbers likely overstate real-world accuracy; cleaner evaluation protocols are needed.

Reading between the lines

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

  • If the decoupling result generalizes, the same recipe—open-vocabulary detector plus frozen LVLM—could be transferred to other disaster-counting tasks (vehicles, shelters, flooded roads) where labeled data are scarce, without retraining the reasoning model.
  • The mask-representation finding implies that for LVLM-based assessment, visual occlusion is more harmful than spatial imprecision; a testable extension is to draw transparent or dashed boxes that preserve more scene context, which may also recover some of the open-source-model losses on FloodNet.
  • The FloodNet annotation mismatch suggests that public benchmark labels encode operational priorities (e.g., access-route flooding) rather than the literal definition; a re-annotation study could quantify how much of the reported error comes from label noise versus model failure.
  • The paper's sequence-leakage caveat implies that reported R² values are optimistic; a strict split by flight mission would give a truer estimate of cross-scene generalization and could change which model combination wins.
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.

Referee Report

4 major / 4 minor

Summary. The paper proposes a two-stage hybrid framework for post-disaster building damage counting from UAV imagery: a Grounding DINO detector generates bounding boxes, and an LVLM performs per-damage-class counting and reasoning. The framework is evaluated on RescueNet and FloodNet with two open-source and two proprietary LVLMs, under three detector fine-tuning regimes (not fine-tuned, FT-100, FT-Full) plus an oracle GTBBox condition. The authors report R², RMSE, and AP75, and include ablations on fine-tuning set size and mask representation, together with qualitative failure analysis and a discussion of dataset limitations. The central claims are that the hybrid framework outperforms isolated baselines, especially under limited detection fine-tuning data, and that decoupling detection from damage assessment is broadly beneficial. Source code and data are publicly released.

Significance. If the hybrid framework were consistently superior to both CV-only and LVLM-only baselines, this would be a practical low-annotation recipe for post-disaster damage counting, a domain where labeled data are scarce and models need to adapt across regions and assessment policies. The paper is also useful for its detailed ablations (fine-tuning set size, mask representation) and its frank acknowledgment of dataset issues such as spatial autocorrelation and label-definition mismatches. The public release of code and data is a strength, and the failure analysis in Section 6.1 is a genuine attempt to explain non-uniform gains. However, the significance is substantially diminished because the headline claim of consistent outperformance is not supported by the paper's own main results table, and the prompt inconsistency in Appendix A.3 raises validity questions for some baselines.

major comments (4)
  1. [§5.3, Table 4] The claim that 'The proposed LVLM+G-DINO framework consistently outperforms all baselines' is contradicted by Table 4. On FloodNet, InternVL 3.5 8B alone achieves R²=0.758, while InternVL+G-DINO FT-100 drops to 0.399, and QwenVL alone (0.529) beats QwenVL+G-DINO FT-100 (0.467). On RescueNet, G-DINO* FT-Full (R²=0.528) outperforms QwenVL+G-DINO FT-Full (-1.374), InternVL+G-DINO FT-Full (-1.028), and even GPT-5.1+G-DINO FT-Full (0.258). The conclusion in Section 7 repeats 'consistently outperforms isolated baselines' without these qualifications. The claims should be restricted to specific model/dataset/fine-tuning combinations where the framework actually wins, or the tables should be honestly presented as showing mixed results.
  2. [§3.2, Table 4] The load-bearing design assumption—that feeding CV-provided bounding boxes improves LVLM damage counting—fails even with perfect boxes for open-source LVLMs on FloodNet. InternVL 3.5 8B + GTBBox gives R²=0.473, which is worse than the InternVL-alone R²=0.758; QwenVL + GTBBox gives 0.421 vs 0.529 alone. This is not a detection-quality issue: the boxes are oracle-perfect. The paper's own Section 6.1 acknowledges degradation, but this admission appears only after the central claim has been stated. The framework's premise is therefore not universal, and the abstract/conclusion need to state the scope of validity explicitly.
  3. [Appendix A.3] The prompt for 'Non-Flooded (without bounding boxes)' instructs: 'Count the number of buildings that are flooded', while the definition describes non-flooded. If this prompt was used for the LVLM-alone baseline on FloodNet, those baseline numbers are invalid because the model was asked to perform the opposite task. The paper must clarify which prompt variant was used for the LVLM-alone baselines, and if the erroneous prompt was used, the FloodNet baseline results and any hybrid-vs-baseline comparisons must be recomputed.
  4. [§5.3, Table 4] No repeated trials, variance estimates, or significance tests are reported. LVLM inference is stochastic (sampling temperature, prompt phrasing, etc.), and several key comparisons are small in magnitude—e.g., QwenVL alone 0.529 vs QwenVL+G-DINO FT-100 0.467, or Gemini 3 Pro alone 0.011 vs Gemini+G-DINO Not FT 0.425. Without multiple runs and confidence intervals, the 'consistently outperforms' language is not statistically grounded. The authors should either provide repeated-run statistics or soften the claims to descriptive observations.
minor comments (4)
  1. [§5.1] The detection confidence threshold (0.3) is a free parameter, but no sensitivity analysis is provided. Given that the paper's conclusions depend on bounding-box quality, a small ablation on this threshold would strengthen the claims.
  2. [§5.4 / Figure 4] The text says 'G-DINO* without any fine-tuning achieves an R² of approximately 0.3' but Table 4 reports R²=0.318 for FloodNet and R²=-3.356 for RescueNet. Please specify that this refers to FloodNet, or clarify the intended claim.
  3. [§6.2] The acknowledged train/val/test overlap from the same flight sequences is a real validity risk, but the paper does not quantify it beyond acknowledging the issue. A simple experiment removing near-duplicate frames from the test set would help gauge contamination magnitude.
  4. [Throughout] Minor language errors appear, e.g., §5.3 'alter models capabilities' and §3.2 'along an confidence scores'. A careful proofread is recommended.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the paper's claims are empirical comparisons on held-out test splits, with no fitted parameter or input definitionally entailing the reported predictions.

full rationale

The paper contains no derivation chain whose conclusion is equivalent to its inputs. The proposed framework is a two-stage pipeline (G-DINO detection followed by LVLM damage counting) evaluated on the held-out test splits of RescueNet and FloodNet. The only trained component, G-DINO, is fine-tuned on FT-100/FT-Full subsets drawn from the training and validation splits, and its performance is then measured on the test sets; this is standard supervised evaluation rather than fitting a parameter to the predicted quantity. LVLMs are deliberately left unfine-tuned, and all reported R^2, RMSE, and AP75 numbers are computed from test-set outputs. No self-citation chain is load-bearing: the paper cites independent prior work (e.g., Grounding DINO, RescueNet, FloodNet, LVLM detection limitations) for component selection, but the central claim of hybrid benefit is supported by the paper's own held-out experiments, not by those citations. The contradictions in Table 4, where open-source LVLMs perform worse with G-DINO boxes than alone, and the prompt-wording inconsistency in Appendix A.3, are validity or soundness concerns about whether the stated conclusion is fully supported; they are not cases where a prediction reduces by construction to a fitted input or to a self-citation. Because no step of the argument equates the output to the input, the circularity score is 0.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

No new entities or mathematically fitted constants. The parameters are experimental choices; the axioms are domain assumptions that the paper itself partially undermines in Section 6.

free parameters (3)
  • G-DINO confidence threshold = 0.3
    Bounding boxes with confidence below 0.3 are discarded (Section 5.1); the threshold is chosen without sensitivity analysis.
  • FT-100 subset size = 100
    Defines the low-data condition (Section 3.3); sampled once with no repeated trials or confidence intervals.
  • LVLM prompt and sampling parameters
    Prompt text is hand-designed (Appendix A.3) and temperature or other decoding parameters for proprietary APIs are not reported.
assumptions (4)
  • domain assumption Bounding boxes from a generic detector improve LVLM damage classification
    Core design premise of Stage 2 (Section 3.2); contradicted by Table 4 for open-source LVLMs on FloodNet.
  • domain assumption Pretrained LVLMs can assess building damage from UAV imagery without fine-tuning
    All LVLMs are used zero-shot (Section 3.3, 'LVLM Fine-tuning').
  • domain assumption Test splits are uncontaminated
    Section 6.2 admits images from the same flight sequence appear across train, validation, and test splits.
  • domain assumption RescueNet and FloodNet annotations are reliable
    Section 6.2 identifies a mismatch between FloodNet's flooded-building definition and its annotations.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Assessing the Benefits of Combining Advanced Deep Learning Techniques for Post-Disaster Building Damage Assessment from UAV Imagery." pith.science (2026). https://pith.science/paper/5IFPA4KC

@misc{pith2026260801906,
  author       = {Pith},
  title        = {Pith review of: Assessing the Benefits of Combining Advanced Deep Learning Techniques for Post-Disaster Building Damage Assessment from UAV Imagery},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5IFPA4KC}},
  note         = {Machine review of arXiv:2608.01906}
}
read the original abstract

Rapid and accurate post-disaster building damage assessment is essential, yet remains a challenging task. Unmanned Aerial Vehicle (UAV) imagery offers a timely and high-resolution view of affected areas, but existing Computer Vision (CV) models often demand large annotated datasets, generalize poorly across geographic regions and their assessment policies, and are confined to the specific tasks they were trained for. Large Vision-Language Models (LVLMs) offer a promising alternative through their strong reasoning and generalization capabilities, but fall short on precise, low-level perception tasks such as object detection and accurate bounding box generation. Furthermore, they often require a substantial amount of data for effective fine-tuning on domain-specific tasks. In this paper, we propose a hybrid framework that decouples detection from damage assessment, combining the precision of CV models with the reasoning power of LVLMs. A CV model first detects buildings and generates bounding boxes on the image that are then passed to an LVLM for damage classification and contextual interpretation. We evaluated our framework on two real-world benchmarks: RescueNet and FloodNet. In particular, the best combination under this framework accurately counts intact, partially damaged and completely destroyed buildings, surpassing isolated baselines by up to 2.1 R^2 points, while requiring only limited annotated data for the detection stage. Beyond reporting aggregate gains, we provide a detailed analysis of failure scenarios and edge cases, offering practical insights for practitioners and concrete directions for future work. Our source code and data are publicly available to the research community via the following repository: https://github.com/ungquanghuy-kddi/VLM_GDINO.git

Figures

Figures reproduced from arXiv: 2608.01906 by the authors.

Figure 1
Figure 1. Overview of our proposed hybrid framework along with its key merits: [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Distribution of average building count per image across RescueNet data [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Distribution of average building count per image across FloodNet data [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figures from the paper (15 more)
Figure 4
Figure 4. Figure 4: Evolution of R2 scores (y-axis) across varying fine-tuning set sizes (x￾axis) of G-DINO* and our proposed LVLM+G-DINO framework on (a) Res￾cueNet and (b) FloodNet. conditions, while consistently outperforming G-DINO*. However, when the fine￾tuning set exceeds 1000 samp…
Figure 5
Figure 5. Figure 5: Impact of mask representation on LVLM damage assessment performance. [PITH_FULL_IMAGE:figures/full_fig_p014_5.png]
Figure 6
Figure 6. Figure 6: Distribution of average pixel occupancy per image across FloodNet data [PITH_FULL_IMAGE:figures/full_fig_p019_6.png]
Figure 7
Figure 7. Figure 7: Distribution of average pixel occupancy per image across RescueNet data [PITH_FULL_IMAGE:figures/full_fig_p020_7.png]
Figure 8
Figure 8. Figure 8: Distribution of samples with existences of flooded and non-flooded build [PITH_FULL_IMAGE:figures/full_fig_p021_8.png]
Figure 9
Figure 9. Figure 9: Scatter plots of actual counts and counts from models on the RescueNet [PITH_FULL_IMAGE:figures/full_fig_p022_9.png]
Figure 10
Figure 10. Figure 10: Scatter plots of actual counts and counts from models on the RescueNet [PITH_FULL_IMAGE:figures/full_fig_p022_10.png]
Figure 11
Figure 11. Figure 11: Scatter plots of actual counts and counts from models on the RescueNet [PITH_FULL_IMAGE:figures/full_fig_p023_11.png]
Figure 12
Figure 12. Figure 12: Scatter plots of actual counts and counts from models on the RescueNet [PITH_FULL_IMAGE:figures/full_fig_p023_12.png]
Figure 13
Figure 13. Figure 13: Scatter plots of actual counts and counts from models on the RescueNet [PITH_FULL_IMAGE:figures/full_fig_p024_13.png]
Figure 14
Figure 14. Figure 14: Scatter plots of actual counts and counts from models on the FloodNet [PITH_FULL_IMAGE:figures/full_fig_p024_14.png]
Figure 15
Figure 15. Figure 15: Scatter plots of actual counts and counts from models on the FloodNet [PITH_FULL_IMAGE:figures/full_fig_p025_15.png]
Figure 16
Figure 16. Figure 16: Scatter plots of actual counts and counts from models on the FloodNet [PITH_FULL_IMAGE:figures/full_fig_p025_16.png]
Figure 17
Figure 17. Figure 17: Scatter plots of actual counts and counts from models on the FloodNet [PITH_FULL_IMAGE:figures/full_fig_p026_17.png]
Figure 18
Figure 18. Figure 18: Scatter plots of actual counts and counts from models on the FloodNet [PITH_FULL_IMAGE:figures/full_fig_p026_18.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

35 extracted references · 9 linked inside Pith

  1. [1]

    org/abs/2402.00157

    Ahn, J., Verma, R., Lou, R., Liu, D., Zhang, R., Yin, W.: Large language models for mathematical reasoning: Progresses and challenges (2024),https://arxiv. org/abs/2402.00157

    arXiv 2024

  2. [2]

    arXiv preprint arXiv:2204.01691 (2022)

    Ahn, M., Brohan, A., Brown, N., Chebotar, Y., Cortes, O., David, B., Finn, C., Fu, C., Gopalakrishnan, K., Hausman, K., et al.: Do as i can, not as i say: Grounding language in robotic affordances. arXiv preprint arXiv:2204.01691 (2022)

    arXiv 2022

  3. [3]

    arXiv preprint arXiv:1906.07155 (2019)

    Chen, K., Wang, J., Pang, J., Cao, Y., Xiong, Y., Li, X., Sun, S., Feng, W., Liu, Z., Xu, J., Zhang, Z., Cheng, D., Zhu, C., Cheng, T., Zhao, Q., Li, B., Lu, X., Zhu, R., Wu, Y., Dai, J., Wang, J., Shi, J., Ouyang, W., Loy, C.C., Lin, D.: MMDetection: Open mmlab detection toolbox and benchmark. arXiv preprint arXiv:1906.07155 (2019)

    arXiv 1906

  4. [4]

    DeepMind, G.: Gemini 3.1 pro (2025),https://deepmind.google/models/ gemini/pro/

    2025

  5. [5]

    Scientific reports12(1) (2022)

    Deng, L., Wang, Y.: Post-disaster building damage assessment based on improved u-net. Scientific reports12(1) (2022)

    2022

  6. [6]

    (2020),https://www.fema

    FEMA: Fema preliminary damage assessment guide. (2020),https://www.fema. gov/sites/default/files/2020-07/fema_preliminary-disaster-assessment_ guide.pdf

    2020

  7. [7]

    Gebre, T.S., Talreja, J., Hashemi-Beni, L.: Multi-Modal Attention for Automated Disaster Damage Assessment Using Remote Sensing Imagery and Deep Learning (2026)

    2026

  8. [8]

    arXiv preprint arXiv:2602.16931 (2026) Post-Disaster Building Damage Assessment from UAV Imagery 17

    Gulati, I., Raval, S.: Narrow fine-tuning erodes safety alignment in vision-language agents. arXiv preprint arXiv:2602.16931 (2026) Post-Disaster Building Damage Assessment from UAV Imagery 17

    arXiv 2026

Show all 35 references
  1. [9]

    In: Proceedings of the IEEE international conference on computer vision

    He, K., Gkioxari, G., Dollár, P., Girshick, R.: Mask r-cnn. In: Proceedings of the IEEE international conference on computer vision. pp. 2961–2969 (2017)

    2017

  2. [10]

    In: ICLR (2022)

    Hu, E.J., Shen, Y., Wallis, P., Allen-Zhu, Z., Li, Y., Wang, S., Wang, L., Chen, W.: Lora: Low-rank adaptation of large language models. In: ICLR (2022)

    2022

  3. [11]

    In: The Thirteenth International Conference on Learning Representations (2025)

    Irvin, J.A., Liu, E.R., Chen, J.C., Dormoy, I., Kim, J., Khanna, S., Zheng, Z., Er- mon, S.: Teochat: A large vision-language assistant for temporal earth observation data. In: The Thirteenth International Conference on Learning Representations (2025)

    2025

  4. [12]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Kuckreja, K., Danish, M.S., Naseer, M., Das, A., Khan, S., Khan, F.S.: Geochat: Grounded large vision-language model for remote sensing. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 27831– 27840 (2024)

    2024

  5. [13]

    arXiv preprint arXiv:2408.03326 (2024)

    Li, B., Zhang, Y., Guo, D., Zhang, R., Li, F., Zhang, H., Zhang, K., Li, Y., Liu, Z., Li, C.: Llava-onevision: Easy visual task transfer. arXiv preprint arXiv:2408.03326 (2024)

    2024 arXiv

  6. [14]

    arXiv preprint arXiv:2404.09610 (2024)

    Lin, Y., Ma, X., Chu, X., Jin, Y., Yang, Z., Wang, Y., Mei, H.: Lora dropout as a sparsity regularizer for overfitting control. arXiv preprint arXiv:2404.09610 (2024)

    2024 arXiv

  7. [15]

    Advances in neural information processing systems36, 34892–34916 (2023)

    Liu, H., Li, C., Wu, Q., Lee, Y.J.: Visual instruction tuning. Advances in neural information processing systems36, 34892–34916 (2023)

    2023

  8. [16]

    In: European conference on computer vision

    Liu, S., Zeng, Z., Ren, T., Li, F., Zhang, H., Yang, J., Jiang, Q., Li, C., Yang, J., Su, H., et al.: Grounding DINO: Marrying DINO with Grounded Pre-training for Open-Set Object Detection. In: European conference on computer vision. pp. 38–55. Springer (2024)

    2024

  9. [17]

    OpenAI: Introducing gpt-5.2 (2025),https://openai.com/index/ introducing-gpt-5-2/

    2025

  10. [18]

    In: Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing

    Pi, R., Gao, J., Diao, S., Pan, R., Dong, H., Zhang, J., Yao, L., Han, J., Xu, H., Kong, L., et al.: Detgpt: Detect what you need via reasoning. In: Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing. pp. 14172–14189 (2023)

    2023

  11. [19]

    Scientific data10(1) (2023)

    Rahnemoonfar, M., Chowdhury, T., Murphy, R.: Rescuenet: A high resolution uav semantic segmentation dataset for natural disaster damage assessment. Scientific data10(1) (2023)

    2023

  12. [20]

    IEEE Access9, 89644–89654 (2021)

    Rahnemoonfar, M., Chowdhury, T., Sarkar, A., Varshney, D., Yari, M., Murphy, R.R.: Floodnet: A high resolution aerial imagery dataset for post flood scene un- derstanding. IEEE Access9, 89644–89654 (2021)

    2021

  13. [21]

    arXiv preprint arXiv:2305.09972 (2023)

    Reis, D., Kupec, J., Hong, J., Daoudi, A.: Real-time flying object detection with yolov8. arXiv preprint arXiv:2305.09972 (2023)

    2023 arXiv

  14. [22]

    In: International Conference on Medical image computing and computer-assisted intervention

    Ronneberger, O., Fischer, P., Brox, T.: U-net: Convolutional networks for biomedi- cal image segmentation. In: International Conference on Medical image computing and computer-assisted intervention. pp. 234–241. Springer (2015)

    2015

  15. [23]

    Frontiers in Earth Science11(2023)

    Sadiq, R., Akhtar, Z., Peterson, S., Keegan, K., El-Sakka, A., Imran, M., Ofli, F.: Towards fine-grained object-level damage assessment during disasters. Frontiers in Earth Science11(2023)

    2023

  16. [24]

    arXiv preprint arXiv:2509.25164 (2025)

    Sapkota, R., Cheppally, R.H., Sharda, A., Karkee, M.: Yolo26: key architectural enhancements and performance benchmarking for real-time object detection. arXiv preprint arXiv:2509.25164 (2025)

    2025

  17. [25]

    Information Fusion126, 103575 (2026)

    Sapkota, R., Karkee, M.: Object detection with multimodal large vision-language models: An in-depth review. Information Fusion126, 103575 (2026)

    2026

  18. [26]

    arXiv preprint arXiv:2508.16016 (2025) 18 Ung et al

    Sirma, A., Plastropoulos, A., Tang, G., Zolotas, A.: Drespnet: A uav dataset and yolov8-drnmodelforaerialinstancesegmentationofbuildingaccesspointsforpost- earthquake search-and-rescue missions. arXiv preprint arXiv:2508.16016 (2025) 18 Ung et al

    2025 arXiv

  19. [27]

    In: Proceedings of the Computer Vision and Pattern Recognition Conference

    Soni, S., Dudhane, A., Debary, H., Fiaz, M., Munir, M.A., Danish, M.S., Fraccaro, P., Watson, C.D., Klein, L.J., Khan, F.S., et al.: Earthdial: Turning multi-sensory earth observations to interactive dialogues. In: Proceedings of the Computer Vision and Pattern Recognition Con...

    2025

  20. [28]

    In: Findings of the Association for Computational Linguistics: ACL 2023

    Suzgun, M., Scales, N., Schärli, N., Gehrmann, S., Tay, Y., Chung, H.W., Chowd- hery, A., Le, Q., Chi, E., Zhou, D., et al.: Challenging big-bench tasks and whether chain-of-thought can solve them. In: Findings of the Association for Computational Linguistics: ACL 2023. pp. 13...

    2023

  21. [29]

    Team, Q.: Qwen3 technical report (2025),https://arxiv.org/abs/2505.09388

    2025 arXiv

  22. [30]

    Wang, A., Liu, L., Chen, H., Lin, Z., Han, J., Ding, G.: Yoloe: Real-time seeing anything.In:ProceedingsoftheIEEE/CVFInternationalConferenceonComputer Vision. pp. 24591–24602 (2025)

    2025

  23. [31]

    In: The Thirty-ninth Annual Conference on Neural Information Processing Systems Datasets and Benchmarks Track (2025)

    Wang, J., Xuan, W., Qi, H., Liu, Z., Liu, K., Wu, Y., Chen, H., Song, J., Xia, J., Zheng, Z., et al.: Disasterm3: A remote sensing vision-language dataset for disaster damage assessment and response. In: The Thirty-ninth Annual Conference on Neural Information Processing Syste...

    2025

  24. [32]

    arXiv preprint arXiv:2508.18265 (2025)

    Wang, W., Gao, Z., Gu, L., Pu, H., Cui, L., Wei, X., Liu, Z., Jing, L., Ye, S., Shao, J., et al.: Internvl3.5: Advancing open-source multimodal models in versatility, reasoning, and efficiency. arXiv preprint arXiv:2508.18265 (2025)

    2025 arXiv

  25. [33]

    IEEE transactions on pattern analysis and machine intelligence45(4), 4768–4781 (2022)

    Weber, E., Papadopoulos, D.P., Lapedriza, A., Ofli, F., Imran, M., Torralba, A.: Incidents1m: a large-scale dataset of images with natural disasters, damage, and incidents. IEEE transactions on pattern analysis and machine intelligence45(4), 4768–4781 (2022)

    2022

  26. [34]

    International Journal of Computer Vi- sion133(2), 825–843 (2025)

    Zang, Y., Li, W., Han, J., Zhou, K., Loy, C.C.: Contextual object detection with multimodal large language models. International Journal of Computer Vi- sion133(2), 825–843 (2025)

    2025

  27. [35]

    In: European Conference on Computer Vision

    Zhan, Y., Zhu, Y., Chen, Z., Yang, F., Tang, M., Wang, J.: Griffon: Spelling out all object locations at any granularity with large language models. In: European Conference on Computer Vision. pp. 405–422. Springer (2024) Post-Disaster Building Damage Assessment from UAV Image...

    2024

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.