Beluga Whale Detection from Satellite Imagery with Point Labels

Guoqing Li; Jinxuan Yang; Yaxuan Wang; Yihang Lu; Yijie Zheng; Yu Chen

REVIEW 4 major objections 5 minor 25 references

Beluga Whale Detection from Satellite Imagery with Point Labels

T0 review · 4 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read This paper claims that an automated point-to-box labeling pipeline using the Segment Anything Model can produce training boxes for beluga whale and harp seal detection that match manually refined boxes, reaching 72.2% F1 for whales…

desk verdict Useful, well-scoped pipeline paper with a real dataset and code, but the headline comparison is less clean than it looks because the ground truth is derived from the very SAM boxes under test. read the letter →

arxiv 2505.12066 v1 pith:4CYZT5KP submitted 2025-05-17 cs.CV

Yijie Zheng , Jinxuan Yang , Yu Chen , Yaxuan Wang , Yihang Lu , Guoqing Li This is my paper · ORCID

classification cs.CV

keywords belugawhaledetectionsatelliteimagerypointannotationsSegmentAnythingModelYOLOv8harpsealmarinemammalmonitoringboundingboxlabeling

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

The paper tries to show that a human only needs to click a point on each animal in very high-resolution satellite imagery; an automated pipeline can then generate the bounding boxes needed to train a detector. It reports that boxes produced by prompting the Segment Anything Model (SAM) with those points outperform fixed-size buffer boxes and nearly match boxes that experts refined by hand, yielding an F1-score of 72.2% for beluga whales overall and 70.3% for harp seals. The study also keeps 'uncertain whales' in the training data, individuals that earlier detection studies tended to discard. If correct, this would cut the annotation labor for marine-mammal monitoring while making counts more complete.

What carries the argument

The load-bearing object is the automated box-labeling pipeline, not any single network. For each manual point annotation, SAM receives the point plus a small buffer box as a prompt and outputs a per-animal mask; in dense scenes, an overlapping-pixel assignment rule gives every contested pixel to the nearest annotation point, producing non-overlapping instance masks that are converted to tight bounding rectangles. These rectangles become the training boxes for YOLOv8s, a single-stage object detector. The pipeline converts cheap point clicks into shape-following boxes, avoiding the fixed-square-box problem of buffer labeling and providing labels for both certain and uncertain whales as well as harp seals.

What would settle it

Re-annotate the test images from scratch with experts who never see SAM output, then retrain YOLO-SAM and YOLO-Buffer and compare on those boxes; if SAM-labeled training no longer beats buffer-labeled training at the reported margins, the central advantage claim would collapse.

Watch

Extended reading notes

Core claim

The central claim is that segmentation-driven annotation from point labels is a viable substitute for manual box drawing in this remote-sensing setting. The study compares three training-label regimes for YOLOv8: fixed-size buffers around each point, raw SAM-generated boxes, and SAM boxes refined by experts. YOLO-SAM reaches the highest whale-overall and harp-seal F1 scores, 72.2% and 70.3%, improves recall for certain whales and harp seals relative to buffer labels, and performs best in dense groups where fixed buffers merge adjacent animals. The paper interprets this as evidence that the annotation bottleneck can be moved from drawing boxes to clicking points, and that SAM-based boxes are precise enough to train an effective detector.

Load-bearing premise

The comparison assumes the expert-refined boxes used as evaluation ground truth are an accurate and independent standard, even though they were created by manually adjusting the same SAM-generated boxes the pipeline produces.

Editorial extensions

If this is right

Manual box annotation can be replaced by point clicking plus an automated segmentation step for beluga and seal monitoring, reducing the human labor bottleneck.
Detection models can be trained to include 'uncertain' whales rather than excluding them, making operational abundance estimates more comprehensive.
Dense aggregations of animals, where fixed buffers merge multiple individuals into one box, become countable with instance-level boxes.
The same point-to-box pipeline should transfer to other species and other very high-resolution imagery sources, since it relies on generic segmentation rather than species-specific features.
The resulting masks can also supply biometric measurements such as body length and width, adding ecological information beyond counts.

Reading between the lines

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

If the pipeline generalizes, historical point-count surveys could be retroactively converted into detection training sets, unlocking decades of archived imagery without new fieldwork.
A natural next test is to apply the pipeline to a different species or sensor without retuning the buffer sizes; success would indicate the method depends on SAM's generalization rather than on dataset-specific tuning.
Because the expert-refined ground truth was derived from SAM boxes, the absolute F1 numbers may be optimistic; the fairest cross-check would be a test set labeled independently from scratch.
The point-to-box idea could be paired with active learning, where SAM boxes flag uncertain animals for human review, to improve label quality exactly where the model struggles.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

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

Worth reading if you care about cutting annotation cost in satellite marine surveys; worth reading with skepticism if you care about benchmarking. The new thing here is a concrete point-to-box pipeline: SAM-H with a point prompt and small buffer, overlapping-pixel nearest-point assignment to separate touching whales, then YOLOv8s trained on the resulting boxes. The dataset is new (538 patches, WorldView-2/3, Clearwater Fiord and Churchill River), the code is public, and the paper includes an honest account of how often expert correction was needed (19% certain whales, 21% uncertain, 4% seals). I checked the numbers in Table II; the headline F1s parse correctly.

The core comparison is YOLO-SAM vs YOLO-Buffer, and SAM boxes win. That result is real-looking, but the stress-test note is right and it matters: the ground truth used to score every model is expert-refined versions of the SAM boxes themselves. The experts corrected only a minority of boxes, so most test-set boxes are close to SAM's output, and YOLO-Buffer's fixed square boxes are scored against a target that was partly built from the method it is being compared against. That shared ancestry can inflate YOLO-SAM's IoU metrics and depress YOLO-Buffer's independent of detection quality. The paper never tests sensitivity to the 0.25 IoU threshold or reports scores against an independent set of boxes drawn without reference to SAM. This is a load-bearing flaw in the comparison, not a quibble.

Other soft spots are smaller. Five runs are averaged but no standard deviations or significance tests appear anywhere. The confidence threshold is tuned on the validation set per model, which is fine but should be disclosed as part of the pipeline. Evaluation is on object-containing patches only, so false alarms over the full 146–1281 km² scenes are not measured. And the abstract's claim that the pipeline 'enhances the detection of uncertain whales' is contradicted by Table II: YOLO-SAM's uncertain-whale F1 is 46.1, lower than YOLO-Buffer's 50.2 and YOLO-Box's 51.3. The authors actually admit this in the body text; the abstract should be aligned with it.

Who is this for? Ecologists and remote-sensing people who want a cheaper annotation route for whale/seal surveys, and anyone building point-to-box label pipelines. The paper deserves serious referee time, but the referee should send it back for an independent ground-truth check or at least an IoU sweep showing the ordering survives, plus error bars and a corrected abstract. As it stands, the pipeline is plausible and usable, but the headline comparison is not fully established.

Referee Report

4 major / 5 minor

Summary. The paper proposes an automated bounding-box labeling pipeline for beluga whale and harp seal detection in very high-resolution satellite imagery. The pipeline uses point annotations together with the Segment Anything Model (SAM) to generate masks that are converted to boxes, and an overlapping-pixel assignment step to separate individuals in dense groups. These SAM-generated boxes are used to train YOLOv8, and the resulting detector is compared against two alternatives: one trained on fixed buffer boxes and one trained on manually refined SAM boxes. On a test set of 538 object-containing patches, the authors report that YOLO-SAM achieves an overall whale F1-score of 72.2% and harp seal F1-score of 70.3%, closely approaching the manually refined YOLO-Box performance and exceeding YOLO-Buffer. The paper argues that the pipeline reduces annotation labor while improving detection, especially in dense scenes, and that the method generalizes to other species and platforms.

Significance. If the central comparison is valid, the paper offers a practical contribution: it shows that point-driven SAM box generation can produce training annotations competitive with (or better than) fixed-size buffers, potentially reducing the manual effort needed for marine mammal monitoring in satellite imagery. The study addresses a concrete annotation bottleneck, and the authors state that the code is publicly available, which supports reproducibility. However, the main comparison is threatened by a structural evaluation bias: the test-set ground truth is built by manually refining the very SAM outputs that the proposed method produces, and the five averaged runs are reported without variance or significance tests. These issues must be resolved before the headline F1 improvements can be accepted as evidence of a real labeling advantage.

major comments (4)

[Section II.B.2 and Table II] The evaluation ground truth for all three YOLO variants is constructed by experts manually refining the SAM-generated masks/boxes, with the paper reporting correction rates of approximately 19% for certain whales, 21% for uncertain whales, and 4% for harp seals. Because YOLO-SAM is trained on the unrefined SAM boxes and evaluated against boxes produced by the same generator, while YOLO-Buffer is evaluated against boxes of a different shape family, the reported F1 ordering (e.g., whale overall 72.2 vs. 63.0) may reflect benchmark construction rather than a genuine labeling-quality improvement. Please provide an independent ground-truth set drawn without reference to SAM outputs (e.g., a fresh expert annotation pass on the test patches), or at minimum an IoU-threshold sweep to demonstrate that the YOLO-SAM versus YOLO-Buffer ordering is robust to the matching criterion.
[Section II.C.3 and Section III.B] The paper states that each experiment was conducted five times and results are reported as averages, but it does not provide standard deviations, confidence intervals, or any significance test. The claim in Section III.B that YOLO-SAM "significantly outperformed" YOLO-Buffer is therefore unsupported. Please report the per-run results or a paired statistical test (e.g., paired bootstrap or McNemar-type test) to establish that the differences are not due to run-to-run variation.
[Section II.C.3] The confidence threshold is tuned on the validation set to maximize the F1-score for certain whales, and the same tuned threshold is then applied to the test set to compare models. This procedure can favor the model whose threshold happens to best fit the validation set, and it does not account for threshold-selection variance. Please report precision-recall curves or results across a range of confidence thresholds to show that the comparative conclusions are not an artifact of the chosen threshold.
[Section II.A.3 and Section II.C.2] The retained dataset includes only image patches containing objects of interest, and the test set is a 20% split of these object-containing patches. Evaluation on such patches inflates precision and recall relative to a realistic full-scene deployment, where most of the water surface contains no animals. Please report results on full scenes or on a data set that includes negative patches, so that false-positive rates in barren areas are captured.

minor comments (5)

[Table II] The column layout of Table II is ambiguous because precision, recall, and F1 values are written as consecutive numbers without clear subheadings or spacing; please reformat the table to make each metric identifiable at a glance.
[Section III.B] The phrase "accuracy across all categories" should be replaced with "precision" or "precision and recall," since the reported metric is precision, not accuracy in the standard classification sense.
[Section II.B.2] The choice of buffer sizes for the SAM prompts (4 meters for whales and 2 meters for harp seals) is stated without justification or sensitivity analysis; a brief rationale or a small ablation would improve the reproducibility.
[Table I] The header "NUMBER OF (SHORT FOR NO.)" is awkward and nonstandard; please replace with a clear column header such as "Number of objects" or "Objects per class."
[References] Reference [22] is marked "in press" without a DOI or year; please update it with complete publication details if available.

Circularity Check

1 steps flagged · score 6.0 of 10

Test-set ground truth is built by expert refinement of the very SAM boxes under test, so the claimed YOLO-SAM advantage is partly circular.

other [Section II.B.2 (labeling procedure) and Section III.A / Table II evaluation]
"The automatically generated bounding boxes were further refined by experts to assess the quality of the SAM annotations. Approximately 19% of certain whale labels, 21% of uncertain whale labels, and 4% of harp seal labels were corrected through manual refinement."

All three detectors are evaluated against test-set boxes that are expert-corrected versions of the SAM-generated boxes (YOLO-SAM's own training labels; for 79-96% of objects, the 'ground truth' is identical to the raw SAM output). YOLO-SAM is therefore scored against a target of the same tight, shape-aligned box family it was trained on, while YOLO-Buffer is scored against boxes that systematically differ from its fixed square outputs. The reported F1 gains (Table II) and the mechanism attributed in Section III.C.1 (shape alignment, reduced overlap) are thus partly forced by the benchmark construction rather than by an independent test of label quality. Expert refinement provides partial independence, so the circularity is not total.

full rationale

The paper's core claim is that SAM-based labels outperform buffer labels for training YOLOv8. The evaluation of that claim uses a test ground truth derived by manually refining the SAM outputs themselves (II.B.2). Since 19-21% of whale labels and 4% of seal labels were corrected, 79-96% of ground-truth boxes coincide with raw SAM boxes, making the comparison between YOLO-SAM and YOLO-Buffer structurally biased: YOLO-SAM is measured against its own output family, while YOLO-Buffer is measured against a shape distribution it was not trained to produce. This is a partial, construction-level circularity, not a complete one, because experts did correct some boxes and the detection F1 still depends on genuine localization ability. The paper is otherwise self-contained: SAM and YOLOv8 are standard external tools, and no load-bearing self-citations or renamed known results appear. The score reflects that the central comparative result is partly an artifact of the ground-truth generation scheme.

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

The claimed F1 scores depend on the buffer sizes, the IoU threshold, and the validation-tuned confidence threshold. The broader pipeline assumes cross-domain generalization of SAM, subjective but expert point labels, and a ground truth that is partially derived from the labeling method being evaluated.

free parameters (3)

SAM prompt buffer size = 4 m for beluga whales, 2 m for harp seals
Hand-chosen input to the prompt box around each point label; directly shapes the SAM-generated masks and boxes (Section II.B.2).
IoU threshold for true positives = 0.25
Chosen to account for small objects; changing it alters all reported F1 values (Section II.C.3).
Confidence threshold per model = not reported
Tuned on the validation set to maximize certain-whale F1; test-set metrics depend on this choice (Section II.C.3).

assumptions (4)

domain assumption SAM generalizes from natural images to segment beluga whales and harp seals in VHR satellite imagery from point prompts.
The pipeline uses SAM-H without any fine-tuning (Section II.B.2); the paper reports only an indirect correction-rate statistic rather than a direct mask-quality validation.
domain assumption Manual point labels reliably locate animals and assign them to certain whale, uncertain whale, or harp seal classes.
Points were placed by interpreters using biometric and spatial context (Section II.A.2); the uncertain class is inherently subjective.
domain assumption Manually refined SAM boxes are an accurate ground truth for evaluating all three labeling strategies.
Expert refinement starts from SAM outputs (Section II.B.2), so the gold standard is not fully independent of the method being tested.
domain assumption Object-containing patches are sufficient for evaluating detection performance.
The dataset kept only the 538 patches containing objects (Section II.A.3); full-scene behavior over empty water is not measured.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Beluga Whale Detection from Satellite Imagery with Point Labels." pith.science (2026). https://pith.science/paper/4CYZT5KP

@misc{pith2026250512066,
  author       = {Pith},
  title        = {Pith review of: Beluga Whale Detection from Satellite Imagery with Point Labels},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4CYZT5KP}},
  note         = {Machine review of arXiv:2505.12066}
}

abstract

Very high-resolution (VHR) satellite imagery has emerged as a powerful tool for monitoring marine animals on a large scale. However, existing deep learning-based whale detection methods usually require manually created, high-quality bounding box annotations, which are labor-intensive to produce. Moreover, existing studies often exclude ``uncertain whales'', individuals that have ambiguous appearances in satellite imagery, limiting the applicability of these models in real-world scenarios. To address these limitations, this study introduces an automated pipeline for detecting beluga whales and harp seals in VHR satellite imagery. The pipeline leverages point annotations and the Segment Anything Model (SAM) to generate precise bounding box annotations, which are used to train YOLOv8 for multiclass detection of certain whales, uncertain whales, and harp seals. Experimental results demonstrated that SAM-generated annotations significantly improved detection performance, achieving higher $\text{F}_\text{1}$-scores compared to traditional buffer-based annotations. YOLOv8 trained on SAM-labeled boxes achieved an overall $\text{F}_\text{1}$-score of 72.2% for whales overall and 70.3% for harp seals, with superior performance in dense scenes. The proposed approach not only reduces the manual effort required for annotation but also enhances the detection of uncertain whales, offering a more comprehensive solution for marine animal monitoring. This method holds great potential for extending to other species, habitats, and remote sensing platforms, as well as for estimating whale biometrics, thereby advancing ecological monitoring and conservation efforts. The codes for our label and detection pipeline are publicly available at http://github.com/voyagerxvoyagerx/beluga-seeker .

Figures

Figures reproduced from arXiv: 2505.12066 by the authors.

**Figure 2.** Automated Box Labeling Procedure. automatically generated bounding boxes were further refined by experts to assess the quality of the SAM annotations. C. Beluga whale detection 1) Whale detection model: YOLO is a single-stage object detection model with proven effectiveness in detecting marine animals in satellite imagery [13, 20]. In this study, we utilized YOLOv8s for its balance of speed and accuracy. YOLOv8s pre… view at source ↗

**Figure 3.** Comparative analysis between SAM-based and buffer box-based annotations (columns 1, 3) and corresponding detection results (columns 2, 4). [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Figures from the paper (1 more)

**Figure 4.** Figure 4: Confusion matrix for beluga whale and harp seal detection. [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

25 extracted references · 20 canonical work pages

[1]

Population dynamics of the threatened Cumberland Sound beluga (Delphinapterus leucas) population,

C. A. Watt, M. Marcoux, S. H. Ferguson, M. O. Hammill, and C. J. Matthews, “Population dynamics of the threatened Cumberland Sound beluga (Delphinapterus leucas) population,” Arctic Science, vol. 7, no. 2, pp. 545–566, 2021. [Online]. Avail- able: https://cdnsciencepub.com/doi/10.1139/as-2019-0030

work page doi:10.1139/as-2019-0030 2021
[2]

Use of satellite imagery to estimate distribution and abundance of Cumberland Sound beluga whales reveals frequent use of a glacial river estuary,

A. M. Belanger, B. A. H. Sherbo, J. D. Roth, and C. A. Watt, “Use of satellite imagery to estimate distribution and abundance of Cumberland Sound beluga whales reveals frequent use of a glacial river estuary,” Frontiers in Marine Science , vol. 10, 2024. [Online]. Available: https: //www.frontiersin.org/articles/10.3389/fmars.2023.1305536

arXiv 2024
[3]

A pioneering survey of deep-diving and off-shore cetaceans in the northern South China Sea,

M. Lin, M. Liu, F. Caruso, M. Rosso, X. Tang, L. Dong, W. Lin, A. Borroni, A. Bocconcelli, L. Dai, and S. Li, “A pioneering survey of deep-diving and off-shore cetaceans in the northern South China Sea,” Integrative Zoology , vol. 16, no. 4, pp. 440–450, 2021. [Online]. Available: https: //onlinelibrary.wiley.com/doi/abs/10.1111/1749-4877.12508

work page arXiv 2021
[4]

C. J. Matthews, C. A. Watt, N. C. Asselin, J. B. Dunn, B. G. Young, L. M. Montsion, K. H. Westdal, P. A. Hall, J. R. Orr, S. H. Ferguson et al., Estimated abundance of the Western Hud- son Bay beluga stock from the 2015 visual and photographic aerial survey. Canadian Science Advisory Secretariat (CSAS), 2017

work page 2015
[5]

The potential of satellite imagery for surveying whales,

C. H ¨oschle, H. C. Cubaynes, P. J. Clarke, G. Humphries, and A. Borowicz, “The potential of satellite imagery for surveying whales,” Sensors, vol. 21, no. 3, p. 963, 2021. [Online]. Available: https://www.mdpi.com/1424-8220/21/3/963

work page 2021
[6]

Annotating very high-resolution satellite imagery: A whale case study,

H. C. Cubaynes, P. J. Clarke, K. T. Goetz, T. Aldrich, P. T. Fretwell, K. E. Leonard, and C. B. Khan, “Annotating very high-resolution satellite imagery: A whale case study,” MethodsX, vol. 10, p. 102040, 2023. [Online]. Available: https: //linkinghub.elsevier.com/retrieve/pii/S2215016123000286

work page 2023
[7]

Eastern High Arctic–Baffin Bay beluga whale ( Delphinapterus leucas) estuary abundance and use from space,

C. A. Watt, C. G. Carlyle, C. A. Hornby, and B. A. H. Sherbo, “Eastern High Arctic–Baffin Bay beluga whale ( Delphinapterus leucas) estuary abundance and use from space,” Frontiers in Mammal Science , vol. 2, p. 1208276, 2023. [Online]. Available: https://www.frontiersin.org/articles/10.3389/fmamm. 2023.1208276/full

work page arXiv 2023
[8]

Using satellite imagery to estimate abundance of Cumberland Sound beluga whales ( Delphinapterus leucas ) in 2021,

B. A. H. Sherbo, A. M. Belanger, B. Charry, and C. A. Watt, “Using satellite imagery to estimate abundance of Cumberland Sound beluga whales ( Delphinapterus leucas ) in 2021,” Frontiers in Marine Science , vol. 11, p. 1422907,

work page 2021

Show all 25 references

[9]

Satellite image survey of beluga whales in the southern Kara Sea,

P. T. Fretwell, H. C. Cubaynes, and O. V . Shpak, “Satellite image survey of beluga whales in the southern Kara Sea,” Marine Mammal Science , vol. 39, no. 4, pp. 1204–1214,
[10]

A semi-automated method using object-based image analysis (OBIA) to detect and enumerate beluga whales in summer from very high-resolution (VHR) satellite imagery,

J. Iacozza, B. Sherbo, and C. Watt, “A semi-automated method using object-based image analysis (OBIA) to detect and enumerate beluga whales in summer from very high-resolution (VHR) satellite imagery,” PLOS ONE , vol. 19, no. 11, p. e0307716, 2024. [Online]. Available: https:/...

2024 doi
[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,” in 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016, pp. 779–788

2016
[12]

Ultralytics YOLO,

G. Jocher, J. Qiu, and A. Chaurasia, “Ultralytics YOLO,” 2023, 8.0.0. [Online]. Available: https://github.com/ultralytics/ ultralytics

2023
[13]

Gray whale detection in satellite imagery using deep learning,

K. M. Green, M. K. Virdee, H. C. Cubaynes, A. I. Aviles- Rivero, P. T. Fretwell, P. C. Gray, D. W. Johnston, C.-B. Sch¨onlieb, L. G. Torres, and J. A. Jackson, “Gray whale detection in satellite imagery using deep learning,” Remote Sensing in Ecology and Conservation , vol. n/...

2023 doi
[14]

Localization and tracking of beluga whales in aerial video using deep learning,

M. Alsaidi, M. G. Al-Jassani, C. Bang, G. O’Corry-Crowe, C. Watt, M. Ghazal, and H. Zhuang, “Localization and tracking of beluga whales in aerial video using deep learning,” Frontiers in Marine Science , vol. 11, p. 1445698,
[15]

Estimating beluga whale abundance from space: using drones to ground-validate VHR satellite imagery,

J. B. Stewart, J. M. Hudson, B. A. H. Sherbo, and C. A. Watt, “Estimating beluga whale abundance from space: using drones to ground-validate VHR satellite imagery,” Remote Sensing in Ecology and Conservation , p. rse2.396, 2024. [Online]. Available: https://zslpublications.onl...

2024 doi
[16]

Mapping arctic cetaceans from space: A case study for beluga and narwhal,

B. Charry, E. Tissier, J. Iacozza, M. Marcoux, and C. A. Watt, “Mapping arctic cetaceans from space: A case study for beluga and narwhal,” PLOS ONE , vol. 16, no. 8, p. e0254380, 2021. [Online]. Available: https://journals.plos.org/plosone/article?id= 10.1371/journal.pone.0254380

2021 doi
[17]

Available: https://www.frontiersin.org/articles/ 10.3389/fmars.2024.1445698/full

[Online]. Available: https://www.frontiersin.org/articles/ 10.3389/fmars.2024.1445698/full

2024
[18]

Segment anything,

A. Kirillov, E. Mintun, N. Ravi, H. Mao, C. Rolland, L. Gustafson, T. Xiao, S. Whitehead, A. C. Berg, W.- Y . Lo, P. Doll ´ar, and R. Girshick, “Segment anything,” 2023, arXiv:2304.02643. [Online]. Available: http://arxiv.org/ abs/2304.02643

2023 arXiv
[19]

SAMRS: Scaling-up remote sensing segmentation dataset with segment anything model,

D. Wang, J. Zhang, B. Du, M. Xu, L. Liu, D. Tao, and L. Zhang, “SAMRS: Scaling-up remote sensing segmentation dataset with segment anything model,” 2023, arXiv:2305.02034. [Online]. Available: http://arxiv.org/abs/2305.02034

2023 arXiv
[20]

Whale detection enhancement through synthetic satellite images,

A. Gaur, C. Liu, X. Lin, N. Karapetyan, and Y . Aloimonos, “Whale detection enhancement through synthetic satellite images,” in OCEANS 2023 - MTS/IEEE U.S. Gulf Coast . IEEE, 2023, pp. 1–7. [Online]. Available: https://ieeexplore. ieee.org/document/10337400/

2023
[21]

Microsoft COCO: Com- mon objects in context,

T.-Y . Lin, M. Maire, S. Belongie, J. Hays, P. Perona, D. Ra- manan, P. Doll ´ar, and C. L. Zitnick, “Microsoft COCO: Com- mon objects in context,” in Computer Vision – ECCV 2014 , D. Fleet, T. Pajdla, B. Schiele, and T. Tuytelaars, Eds. Cham: Springer International Publishing...

2014
[22]

Deep learning for automated shark detection and biometrics without keypoints,

J. Clark, C. Lalgudi, M. Leone, J. Meribe, and M. Espinoza, “Deep learning for automated shark detection and biometrics without keypoints,” in Computer Vision – ECCV 2024 Work- shops, in press

2024
[23]

Deep learning based whale detection from satellite imagery,

S. Kapoor, M. Kumar, and M. Kaushal, “Deep learning based whale detection from satellite imagery,” Sustainable Computing: Informatics and Systems , vol. 38, p. 100858, 2023. [Online]. Available: https://linkinghub.elsevier.com/retrieve/pii/ S2210537923000136

2023
[2023]

Available: https://onlinelibrary.wiley.com/doi/ 10.1111/mms.13044

[Online]. Available: https://onlinelibrary.wiley.com/doi/ 10.1111/mms.13044

doi
[2024]

Available: https://www.frontiersin.org/articles/ 10.3389/fmars.2024.1422907/full

[Online]. Available: https://www.frontiersin.org/articles/ 10.3389/fmars.2024.1422907/full

2024

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.

[1] [1]

Population dynamics of the threatened Cumberland Sound beluga (Delphinapterus leucas) population,

C. A. Watt, M. Marcoux, S. H. Ferguson, M. O. Hammill, and C. J. Matthews, “Population dynamics of the threatened Cumberland Sound beluga (Delphinapterus leucas) population,” Arctic Science, vol. 7, no. 2, pp. 545–566, 2021. [Online]. Avail- able: https://cdnsciencepub.com/doi/10.1139/as-2019-0030

work page doi:10.1139/as-2019-0030 2021

[2] [2]

Use of satellite imagery to estimate distribution and abundance of Cumberland Sound beluga whales reveals frequent use of a glacial river estuary,

A. M. Belanger, B. A. H. Sherbo, J. D. Roth, and C. A. Watt, “Use of satellite imagery to estimate distribution and abundance of Cumberland Sound beluga whales reveals frequent use of a glacial river estuary,” Frontiers in Marine Science , vol. 10, 2024. [Online]. Available: https: //www.frontiersin.org/articles/10.3389/fmars.2023.1305536

arXiv 2024

[3] [3]

A pioneering survey of deep-diving and off-shore cetaceans in the northern South China Sea,

M. Lin, M. Liu, F. Caruso, M. Rosso, X. Tang, L. Dong, W. Lin, A. Borroni, A. Bocconcelli, L. Dai, and S. Li, “A pioneering survey of deep-diving and off-shore cetaceans in the northern South China Sea,” Integrative Zoology , vol. 16, no. 4, pp. 440–450, 2021. [Online]. Available: https: //onlinelibrary.wiley.com/doi/abs/10.1111/1749-4877.12508

work page arXiv 2021

[4] [4]

C. J. Matthews, C. A. Watt, N. C. Asselin, J. B. Dunn, B. G. Young, L. M. Montsion, K. H. Westdal, P. A. Hall, J. R. Orr, S. H. Ferguson et al., Estimated abundance of the Western Hud- son Bay beluga stock from the 2015 visual and photographic aerial survey. Canadian Science Advisory Secretariat (CSAS), 2017

work page 2015

[5] [5]

The potential of satellite imagery for surveying whales,

C. H ¨oschle, H. C. Cubaynes, P. J. Clarke, G. Humphries, and A. Borowicz, “The potential of satellite imagery for surveying whales,” Sensors, vol. 21, no. 3, p. 963, 2021. [Online]. Available: https://www.mdpi.com/1424-8220/21/3/963

work page 2021

[6] [6]

Annotating very high-resolution satellite imagery: A whale case study,

H. C. Cubaynes, P. J. Clarke, K. T. Goetz, T. Aldrich, P. T. Fretwell, K. E. Leonard, and C. B. Khan, “Annotating very high-resolution satellite imagery: A whale case study,” MethodsX, vol. 10, p. 102040, 2023. [Online]. Available: https: //linkinghub.elsevier.com/retrieve/pii/S2215016123000286

work page 2023

[7] [7]

Eastern High Arctic–Baffin Bay beluga whale ( Delphinapterus leucas) estuary abundance and use from space,

C. A. Watt, C. G. Carlyle, C. A. Hornby, and B. A. H. Sherbo, “Eastern High Arctic–Baffin Bay beluga whale ( Delphinapterus leucas) estuary abundance and use from space,” Frontiers in Mammal Science , vol. 2, p. 1208276, 2023. [Online]. Available: https://www.frontiersin.org/articles/10.3389/fmamm. 2023.1208276/full

work page arXiv 2023

[8] [8]

Using satellite imagery to estimate abundance of Cumberland Sound beluga whales ( Delphinapterus leucas ) in 2021,

B. A. H. Sherbo, A. M. Belanger, B. Charry, and C. A. Watt, “Using satellite imagery to estimate abundance of Cumberland Sound beluga whales ( Delphinapterus leucas ) in 2021,” Frontiers in Marine Science , vol. 11, p. 1422907,

work page 2021