Pith. sign in

REVIEW 4 major objections 6 minor 30 references

AI-Driven Real-Time Monitoring of Ground-Nesting Birds: A Case Study on Curlew Detection Using YOLOv10

T0 review · 4 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read A real-time camera-trap system using YOLOv10 detects adult curlews and chicks with 90–96% F1 scores in a Welsh field trial.

desk verdict A genuinely deployed YOLOv10 curlew detector with plausible training metrics, but the field-performance claims rely on an undocumented labeling protocol and threshold; a conditional accept after revision. read the letter →

arxiv 2411.15263 v1 pith:BR3AJXPD submitted 2024-11-22 cs.CV cs.AI

Carl Chalmers , Paul Fergus , Serge Wich , Steven N Longmore , Naomi Davies Walsh , Lee Oliver , James Warrington , Julieanne Quinlan
show 1 more author
Katie Appleby
This is my paper · ORCID
classification cs.CVcs.AI
keywords conservationobjectdetectionimageprocessingmodellingbiodiversityDeepLearningcameratrapsYOLOv10curlew
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 tries to show that AI object detection can move curlew monitoring from manual, delayed camera-trap review to real-time alerts. It reports a custom-trained YOLOv10 model, integrated with the Conservation AI platform and 3/4G-enabled cameras, that classifies adult curlews and chicks as images arrive. Across 11 nesting sites in Wales over about six weeks, the model achieved sensitivity of 90.56% for adults and 92.35% for chicks, with 100% specificity and F1 scores of 95.05% and 96.03% respectively. If these numbers hold, conservationists could receive immediate alerts about nesting activity and chick presence, allowing faster intervention for a species in steep decline.

What carries the argument

The load-bearing mechanism is YOLOv10x, a single-stage, anchor-free object detector that predicts bounding boxes and class probabilities in one pass using a CSPDarknet backbone and a Path Aggregation Network for multi-scale feature fusion. The model was pre-trained on MS COCO and fine-tuned via transfer learning on 38,740 tagged images spanning 26 UK species and objects, then exported to ONNX and served behind a GPU inference server so that camera-trap images uploaded over 3/4G are classified in real time. This combination of single-stage detection, transfer learning, and platform integration is what lets the system deliver species-level classifications without manual triage.

What would settle it

Have independent experts re-annotate all 1,072 trial images without seeing the model's outputs, then recompute the confusion matrices for adult curlews and chicks; if the expert labels add missed individuals or false-positive background objects, the reported sensitivities (90.56%, 92.35%) and 100% specificity will not reproduce.

Watch

Extended reading notes

Core claim

The central claim is that a single-stage YOLOv10x detector, fine-tuned on a 26-class UK species dataset, can be embedded in a real-time camera-trap pipeline and reliably detect and classify Eurasian curlews (Numenius arquata) and their chicks under field conditions. The model processes images transmitted by 3/4G cellular cameras through the Conservation AI platform; during the trial, 1,072 images from 11 Welsh nesting sites were classified automatically. The paper reports per-class inference metrics of 93.41% accuracy, 100% precision, 90.56% sensitivity, 100% specificity, and a 95.05% F1 score for adult curlews, and 97.51% accuracy, 100% precision, 92.35% sensitivity, 100% specificity, and a 96.03% F1 score for chicks, with domestic sheep also detected at 100% across all metrics. It also reports that the system filtered irrelevant images with 98.28% accuracy, reducing the manual review burden.

Load-bearing premise

The field-performance numbers assume that the 1,072 deployment images were labelled with accurate ground truth independently of what the model predicted, but the paper does not describe who created those labels, how they were verified, or how true negatives and missed detections were counted.

Editorial extensions

If this is right

  • If the reported performance holds, curlew conservation teams can receive near-real-time alerts when an adult or chick appears at a nest, enabling faster anti-predator or habitat interventions.
  • Automated filtering of blank and irrelevant images (98.28% accurate in the trial) cuts the manual-review workload that currently delays camera-trap analysis.
  • The same 26-class model and pipeline can be extended to monitor other ground-nesting birds and mammals without retraining the full system from scratch.
  • The authors state that the deployment provides a platform for a longitudinal curlew nesting-season survey in 2025, which would test whether the detection metrics translate into measurable conservation outcomes.

Reading between the lines

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

  • Because the trial ran for roughly six weeks at 11 sites in one region, the reported 100% specificity and high sensitivities are estimates for that deployment window; broader seasons and habitats could introduce new false positives or missed chicks that the current numbers do not capture.
  • The confusion between adult curlews and common pheasants suggests that visually similar ground-nesting birds may need class-specific training data or a hierarchical classifier before the system can be trusted for multi-species monitoring.
  • The same real-time alert architecture could be pointed at predators such as foxes, badgers, or corvids, turning a detection system into an early-warning system for predation risk rather than only a presence/absence logger.
  • If paired with standardized camera-placement guidelines, citizen-deployed cameras could scale this approach across the curlew's range; the paper itself notes that camera placement strongly affected chick detections.
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 / 6 minor

Summary. The paper reports a real-time camera-trap monitoring pipeline for Eurasian curlews (Numenius arquata) and their chicks, built around a custom fine-tuned YOLOv10x detector integrated with the Conservation AI platform. The model was trained on 38,740 images spanning 26 UK species/objects and reports a held-out test mAP of 0.976. The deployment study at 11 nesting sites in Wales over May–June 2024 analyzed 1,072 images, reporting for adult curlews a sensitivity of 90.56%, specificity of 100%, and F1-score of 95.05%, and for chicks a sensitivity of 92.35%, specificity of 100%, and F1-score of 96.03%. The paper claims the system provides timely, scalable conservation monitoring, with the main contribution being the integration of a high-accuracy detector into a real-time pipeline and its field evaluation.

Significance. If the field-performance claims are correct, the work is practically valuable: it would demonstrate that a deployed YOLOv10-based pipeline can detect adult curlews and chicks in real-world camera-trap imagery with high sensitivity and no false positives over a six-week trial, which is directly relevant to curlew conservation and similar ground-nesting bird monitoring. The training component is standard but solid, with a respectable mAP and clear reporting of hyperparameters and augmentation settings. The main significance hinges on the auditability of the deployment evaluation; as presented, the central field metrics rest on an undocumented ground-truth process for the 1,072 inference images, and on unstated decisions about confidence thresholds and metric aggregation. The paper does not release data or code, which further limits independent verification, though the protocol itself could be clarified in a revision.

major comments (4)
  1. [Section 3.2 and Table 1] The ground-truth labeling of the 1,072 deployment images is not described. The paper specifies how training data were tagged (Section 2.2) but is silent on who labeled the field images, whether labels were independently verified, how false negatives were found (e.g., whether every image was reviewed by a human), and how true negatives were defined at image level versus object level. The acknowledgment that Rachel Chalmers tagged 'all the data' does not clarify her role in the deployment evaluation. Without this protocol, the reported sensitivity and specificity for curlews and chicks in Table 1 cannot be audited, so this is a load-bearing omission for the paper's central claim.
  2. [Section 2.7/Figure 12 and Section 3.2.1/Table 1] The inference confidence threshold used to produce Table 1 is never reported. The F1-confidence curve in Figure 12 shows a peak at a confidence threshold of 0.387, but the text never states that this threshold (or any other) was applied during the deployment. Since precision, recall, and specificity are threshold-dependent, the reader cannot reproduce the reported metrics or assess whether the chosen threshold was selected post hoc. Please state the exact threshold used and, if possible, report metrics across a range of thresholds for the deployment data.
  3. [Section 3.2.1, Table 1, and Table 2] The reported overall accuracy (91.23%) and average F1-score (58.88%) are inconsistent with the three-class Table 1. The text says 'individual class accuracies ranging from 93.41% to 100% and an overall accuracy of 91.23% (Table 1)', but the average of the three displayed accuracies is approximately 96.97%, not 91.23%. The paper also states that Common pheasant had zero true instances yet contributed false positives, and that some classes were 'discontinued from the analysis'. Table 2, which should provide the full confusion matrix, appears empty or incomplete in the manuscript. Please clarify which classes were included in the averaged metrics, how accuracy was aggregated (micro vs. macro average, image-level vs. object-level), and provide the complete confusion matrix with counts for all classes.
  4. [Discussion (Section 4) and Section 3.2] The paper acknowledges in the Discussion that 'Not all camera trap installations in the study adhered to these guidelines, consequently some misdetections were observed' regarding camera placement, yet the quantitative impact of these misdetections is not reflected in the reported metrics. This is not necessarily an error, but it raises a question about whether Table 1's sensitivity values include all deployment images or only a subset from well-placed cameras. Please specify whether any images or sites were excluded from the evaluation, and if so, how the exclusion decision was made.
minor comments (6)
  1. [Section 2.6] The paragraph beginning 'The end-to-end inferencing pipeline as shown in Figure 6...' is repeated verbatim within the same subsection; one copy should be deleted.
  2. [Section 2] Section numbering is inconsistent: what appears to be Section 2.1 is labeled '3.1. Data Collection and Description', and Section 2.7 is labeled '3.8. Evaluation Metrics Inference'. Please renumber all sections consistently.
  3. [Section 2.2 and Section 2.4] The text says 'The dataset used in this study comprised a total of 38,740 image files' but later says 'In total, 38,740 objects were tagged across the dataset.' These are different quantities; please clarify whether 38,740 refers to images, annotated objects, or both.
  4. [Abstract and Section 4] The claim that the system 'filter[s] out blank images triggered by moving vegetation' with an accuracy of 98.28% appears only in the Discussion and is not supported by any results section or table; please either provide the supporting data or remove the specific number.
  5. [Throughout] There are multiple typographical errors, including 'du e' in the abstract and 'Northan goshawk' in the species list. Please perform a careful proofreading pass.
  6. [Table 2] Table 2's caption says 'The diagonal number indicates the TP for each of the classes', but the actual matrix contents are not visible in the manuscript. Please include the full matrix with row and column labels.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the reported metrics are measured outcomes of a supervised detector, not derived from the model's own outputs by construction.

full rationale

The paper's claimed derivation is a standard supervised training and evaluation chain. The YOLOv10x model is trained on 38,740 hand-tagged images split into 80/10/10 sets, test-split performance (mAP 0.976 at IoU 0.5) is measured on held-out images, and the deployment metrics in Table 1 are computed from TP/FP/TN/FN counts on 1,072 trial images using the standard definitions in Eqs. 7-10. No fitted parameter is renamed as a prediction: the reported F1 values are recomputed from the reported precision and sensitivity values, and specificity reaches 100% because no false positives were recorded for those classes. The self-citations in Refs. 14 and 26 describe the authors' prior bird-classification work and the Conservation AI platform, but they do not carry the load of the central performance claim; no uniqueness theorem or ansatz is imported from those works. The paper's genuine weakness is auditability, not circularity: the labeling and verification protocol for the 1,072 inference images and the confidence threshold used at deployment are not reported, and the reported overall accuracy (91.23%) and average F1 (58.88%) imply unlisted classes beyond Table 1. These are correctness and reproducibility concerns, not evidence that the results reduce by construction to their inputs. I therefore find no significant circularity.

Assumptions & free parameters 7 free parameters · 4 assumptions · 0 invented entities

The central claim rests on annotation correctness, independent field labels, transfer-learning assumptions, and a representative deployment. None of these are independently verified in the paper, and no data or code artifacts are released.

free parameters (7)
  • YOLOv10x learned weights = not released
    All detection outputs and reported metrics depend on millions of weights fit to the training data; the weights are not provided.
  • Inference confidence threshold = not reported; F1-Confidence curve peaks at 0.387 on validation
    Precision, recall, and F1 are threshold-dependent, but the threshold used during the 1,072-image deployment is not stated.
  • Learning rate = 0.01
    Chosen for training; affects convergence and final model behavior.
  • Batch size = 256
    Chosen for training stability and GPU memory constraints.
  • Image size = 640 pixels
    Chosen to balance detection accuracy and computational efficiency.
  • Momentum = 0.937
    Chosen for optimizer stability during gradient descent.
  • Augmentation coefficients = hsv_h=0.015, hsv_s=0.7, hsv_v=0.4, fliplr=0.5, translate=0.1, scale=0.5, erasing=0.4
    Chosen to diversify training samples; these settings influence generalization and final metrics.
assumptions (4)
  • domain assumption The 38,740 training images were annotated with correct bounding boxes and class labels.
    Every training and test metric inherits the accuracy of the annotations. A single tagger is acknowledged, and no inter-annotator agreement or independent verification is reported.
  • domain assumption The 1,072 inference images received accurate ground-truth labels independent of the model's predictions.
    Table 1 claims sensitivity, specificity, and F1 for the deployment, but the paper never describes the labeling or verification procedure for these field images.
  • domain assumption Pre-training on MS COCO provides a useful feature initialization for camera-trap wildlife detection after fine-tuning.
    The transfer-learning argument in Section 2.4 assumes that COCO features transfer to ecologically similar species in camera-trap images.
  • domain assumption The random 80/10/10 split at image level prevents data leakage and the test set represents the deployment distribution.
    The split is described in Section 3.1, but the paper does not check whether images from the same camera or event appear across splits, or whether deployment conditions match training conditions.

how reviews work

0 comments
Cite this review

Pith. "Pith review of AI-Driven Real-Time Monitoring of Ground-Nesting Birds: A Case Study on Curlew Detection Using YOLOv10." pith.science (2026). https://pith.science/paper/BR3AJXPD

@misc{pith2026241115263,
  author       = {Pith},
  title        = {Pith review of: AI-Driven Real-Time Monitoring of Ground-Nesting Birds: A Case Study on Curlew Detection Using YOLOv10},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BR3AJXPD}},
  note         = {Machine review of arXiv:2411.15263}
}
read the original abstract

Effective monitoring of wildlife is critical for assessing biodiversity and ecosystem health, as declines in key species often signal significant environmental changes. Birds, particularly ground-nesting species, serve as important ecological indicators due to their sensitivity to environmental pressures. Camera traps have become indispensable tools for monitoring nesting bird populations, enabling data collection across diverse habitats. However, the manual processing and analysis of such data are resource-intensive, often delaying the delivery of actionable conservation insights. This study presents an AI-driven approach for real-time species detection, focusing on the curlew (Numenius arquata), a ground-nesting bird experiencing significant population declines. A custom-trained YOLOv10 model was developed to detect and classify curlews and their chicks using 3/4G-enabled cameras linked to the Conservation AI platform. The system processes camera trap data in real-time, significantly enhancing monitoring efficiency. Across 11 nesting sites in Wales, the model achieved high performance, with a sensitivity of 90.56%, specificity of 100%, and F1-score of 95.05% for curlew detections, and a sensitivity of 92.35%, specificity of 100%, and F1-score of 96.03% for curlew chick detections. These results demonstrate the capability of AI-driven monitoring systems to deliver accurate, timely data for biodiversity assessments, facilitating early conservation interventions and advancing the use of technology in ecological research.

Figures

Figures reproduced from arXiv: 2411.15263 by the authors.

Figure 1
Figure 1. Camera trap image of a Curlew visiting one of areas used in the study. The remainder of this paper is structured as follows: Section 2 outlines the methodology, including data collection, model training, and evaluation metrics. Section 3 presents the results, including both model performance and real-time inference for the curlew case study, followed by the discussion in Section 4. The conclusions and future directi… view at source ↗
Figure 2
Figure 2. As a result, no images were excluded [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 2
Figure 2. Distribution of image resolutions across the dataset, highlighting any outliers that may impact model training and performance. 2.2. Data Pre-Processing The data tagging process was conducted using the Conservation AI tagging platform, where bounding boxes were applied to delineate regions of interest within each image. These tagged regions were exported in Extensible Markup Language (XML) format, adhering to the Pa… view at source ↗
Figures from the paper (13 more)
Figure 3
Figure 3. Figure 3: Class distribution of annotated data showing the species name on the x axis and the number of tags per species on the y axis. The labeled data was converted to the YOLO annotation format using a Python script. The dataset, consisting of images and corresponding labels,…
Figure 4
Figure 4. Figure 4: YOLOv10 architecture used in the study. 2.4. Transfer Learning Transfer learning enables the adaptation of a pre-trained model to new tasks by fine-tuning its learned parameters for novel objects or species of interest [28]. This technique is critical when working with…
Figure 5
Figure 5. Figure 5: 3/4G camera trap used for real-time inference. Throughout the study the camera trap was continuously charged using a solar panel. The end-to-end inferencing pipeline as shown in [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: End-to-end inferencing pipeline for the Conservation AI platform [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: Conservation AI inferencing platform using Triton server. The end-to-end inferencing pipeline as shown in [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 8
Figure 8. Figure 8: Inference processing pipeline. 2.7. Evaluation Metrics Training The model is evaluated using the test split following training to assess its generalisation performance before deployment in real-time systems. This evaluation provides a comprehensive view of the model's …
Figure 9
Figure 9. Figure 9: Precision-Recall (PR) curve for the model trained on UK species and objects. The Precision-Confidence curve for the model, as illustrated in [PITH_FULL_IMAGE:figures/full_fig_p010_9.png]
Figure 10
Figure 10. Figure 10: Precision-Confidence curve for the model trained on UK species and objects. The Recall-Confidence curve, as presented in [PITH_FULL_IMAGE:figures/full_fig_p011_10.png]
Figure 11
Figure 11. Figure 11: Recall-Confidence curve for the model trained on UK species and objects [PITH_FULL_IMAGE:figures/full_fig_p012_11.png]
Figure 13
Figure 13. Figure 13: Confusion matrix for the model trained on UK species and objects. 3.2. Model Deployment The trained model was deployed to evaluate its performance during the trial period. Using the inferencing pipeline and camera setup described in the methodology section, the system…
Figure 14
Figure 14. Figure 14: Example detections of Common curlew (adult) during the inference trial using one of the 3/4G cameras. As illustrated in [PITH_FULL_IMAGE:figures/full_fig_p016_14.png]
Figure 15
Figure 15. Figure 15: Example detection of a Common curlew chick during the trial using one of the 3/4G cameras [PITH_FULL_IMAGE:figures/full_fig_p017_15.png]
Figure 16
Figure 16. Figure 16: Example detection of a curlew and curlew chick obtained from a real-time camera. Despite these achievements, several issues were identified. Misclassifications, particularly between common pheasants and curlews were notable during the study. Furthermore, correct camer…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

30 extracted references · 24 canonical work pages

  1. [1]

    State of nature 2023,

    F. Burns et al., “State of nature 2023,” 2023

    work page 2023

  2. [2]

    Reducing nest predation of ground -nesting birds through conditioned food aversion,

    J. Tobajas, E. Descalzo, R. Mateo, and P. Ferreras, “Reducing nest predation of ground -nesting birds through conditioned food aversion,” Biol Conserv , vol. 242, p. 108405, Feb. 2020, doi: 10.1016/j.biocon.2020.108405

    work page arXiv 2020

  3. [3]

    Revealing the contributions of reproduction and survival to the Europe-wide decline in meadow birds: Review and meta-analysis,

    M. Roodbergen, B. van der Werf, and H. Hoetker, “Revealing the contributions of reproduction and survival to the Europe-wide decline in meadow birds: Review and meta-analysis,” J Ornithol, vol. 153, no. 1, pp. 53–74, Jan. 2012, doi: 10.1007/s10336-011-0733-y. 19

    work page doi:10.1007/s10336-011-0733-y 2012

  4. [4]

    Revisiting an old question: Which predators eat eggs of ground-nesting birds in farmland landscapes?,

    C. Bravo, O. Pays, M. Sarasa, and V. Bretagnolle, “Revisiting an old question: Which predators eat eggs of ground-nesting birds in farmland landscapes?,” Science of The Total Environment, vol. 744, p. 140895, Nov. 2020, doi: 10.1016/j.scitotenv.2020.140895

    work page arXiv 2020

  5. [5]

    Ecosystem Services Provided by Birds,

    C. J. Whelan, D. G. Wenny, and R. J. Marquis, “Ecosystem Services Provided by Birds,” Ann N Y Acad Sci, vol. 1134, no. 1, pp. 25–60, Jun. 2008, doi: 10.1196/annals.1439.003

    work page doi:10.1196/annals.1439.003 2008

  6. [6]

    Bird - mediated effects of pest control services on crop productivity: a global synthesis,

    P. Díaz -Siefer, N. Olmos -Moya, F. E. Fontúrbel, B. Lavandero, R. A. Pozo, and J. L. Celis -Diez, “Bird - mediated effects of pest control services on crop productivity: a global synthesis,” J Pest Sci (2004), vol. 95, no. 2, pp. 567–576, Mar. 2022, doi: 10.1007/s10340-021-01438-4

    work page doi:10.1007/s10340-021-01438-4 2004

  7. [7]

    Home ranges and hatching success of threatened Eurasian curlew in north-eastern Europe relates to habitat type: Natural vs. agricultural landscapes,

    P. Bocher et al., “Home ranges and hatching success of threatened Eurasian curlew in north-eastern Europe relates to habitat type: Natural vs. agricultural landscapes,” Glob Ecol Conserv, vol. 50, p. e02851, 2024, doi: https://doi.org/10.1016/j.gecco.2024.e02851

    work page doi:10.1016/j.gecco.2024.e02851 2024

  8. [8]

    A synthesis of Eurasian Curlew ( Numenius arquata arquata ) demography and population viability to inform its management,

    D. S. Viana, S. Santoro, R. C. Soriguer, and J. Figuerola, “A synthesis of Eurasian Curlew ( Numenius arquata arquata ) demography and population viability to inform its management,” Ibis, vol. 165, no. 3, pp. 767–780, Jul. 2023, doi: 10.1111/ibi.13184

    work page doi:10.1111/ibi.13184 2023

Show all 30 references
  1. [9]

    Nest survival of threatened Eurasian Curlew ( Numenius arquata ) breeding at low densities across a human‐modified landscape,

    H. Ewing, S. Franks, J. Smart, N. Burton, and J. A. Gill, “Nest survival of threatened Eurasian Curlew ( Numenius arquata ) breeding at low densities across a human‐modified landscape,” Ibis, vol. 165, no. 3, pp. 753–766, Jul. 2023, doi: 10.1111/ibi.13180

    2023 doi

  2. [10]

    Protecting wild bird nests against predators: A systematic review and meta‐analysis of non‐lethal methods,

    D. Gautschi, A. Čulina, R. Heinsohn, D. Stojanovic, and R. Crates, “Protecting wild bird nests against predators: A systematic review and meta‐analysis of non‐lethal methods,” Journal of Applied Ecology , vol. 61, no. 6, pp. 1187–1198, Jun. 2024, doi: 10.1111/1365-2664.14619

    2024

  3. [11]

    Camera Traps in Animal Ecology,

    T. O’Brien, “Camera Traps in Animal Ecology,” 2011, pp. 71–96. doi: 10.1007/978-4-431-99495-4_6

    2011 doi

  4. [12]

    Snap happy: camera traps are an effective sampling tool when compared with alternative methods,

    O. R. Wearn and P. Glover -Kapfer, “Snap happy: camera traps are an effective sampling tool when compared with alternative methods,” R Soc Open Sci, vol. 6, no. 3, p. 181748, 2019, doi: 10.1098/rsos.181748

    2019 doi

  5. [13]

    Software to facilitate and streamline camera trap data management: A review,

    S. Young, J. Rode‐Margono, and R. Amin, “Software to facilitate and streamline camera trap data management: A review,” Ecol Evol, vol. 8, no. 19, pp. 9947–9957, Oct. 2018, doi: 10.1002/ece3.4464

    2018 doi

  6. [14]

    Removing Human Bottlenecks in Bird Classification Using Camera Trap Images and Deep Learning,

    C. Chalmers et al., “Removing Human Bottlenecks in Bird Classification Using Camera Trap Images and Deep Learning,” Remote Sens (Basel), vol. 15, no. 10, p. 2638, May 2023, doi: 10.3390/rs15102638

    2023 doi

  7. [15]

    An evaluation of platforms for processing camera -trap data using artificial intelligence,

    J. Vélez et al., “An evaluation of platforms for processing camera -trap data using artificial intelligence,” Methods Ecol Evol, vol. 14, no. 2, pp. 459–477, 2023, doi: https://doi.org/10.1111/2041-210X.14044

    2023 doi

  8. [16]

    Automatically identifying, counting, and describing wild animals in camera - trap images with deep learning,

    M. S. Norouzzadeh et al., “Automatically identifying, counting, and describing wild animals in camera - trap images with deep learning,” Proceedings of the National Academy of Sciences, vol. 115, no. 25, pp. E5716– E5725, 2018

    2018

  9. [17]

    Scaling‐up camera traps: monitoring the planet’s biodiversity with networks of remote sensors,

    R. Steenweg et al., “Scaling‐up camera traps: monitoring the planet’s biodiversity with networks of remote sensors,” Front Ecol Environ, vol. 15, no. 1, pp. 26–34, Feb. 2017, doi: 10.1002/fee.1448

    2017 doi

  10. [18]

    Efficient pipeline for camera trap image review,

    S. Beery, D. Morris, and S. Yang, “Efficient pipeline for camera trap image review,” arXiv preprint arXiv:1907.06772, 2019

    1907 arXiv

  11. [19]

    Pytorch -Wildlife: A Collaborative Deep Learning Framework for Conservation,

    A. Hernandez, Z. Miao, L. Vargas, R. Dodhia, and J. Lavista, “Pytorch -Wildlife: A Collaborative Deep Learning Framework for Conservation,” arXiv preprint arXiv:2405.12930, 2024

    2024 arXiv

  12. [20]

    Wildlife Insights: A Platform to Maximize the Potential of Camera Trap and Other Passive Sensor Wildlife Data for the Planet,

    J. A. Ahumada et al., “Wildlife Insights: A Platform to Maximize the Potential of Camera Trap and Other Passive Sensor Wildlife Data for the Planet,” Environ Conserv , vol. 47, no. 1, pp. 1 –6, Mar. 2020, doi: 10.1017/S0376892919000298

    2020 doi

  13. [21]

    The inaturalist species classification and detection dataset,

    G. Van Horn et al., “The inaturalist species classification and detection dataset,” in Proceedings of the IEEE conference on computer vision and pattern recognition, 2018, pp. 8769–8778

    2018

  14. [22]

    Snapshot Serengeti, high - frequency annotated camera trap images of 40 mammalian species in an African savanna,

    A. Swanson, M. Kosmala, C. Lintott, R. Simpson, A. Smith, and C. Packer, “Snapshot Serengeti, high - frequency annotated camera trap images of 40 mammalian species in an African savanna,” Sci Data, vol. 2, no. 1, p. 150026, Jun. 2015, doi: 10.1038/sdata.2015.26

    2015 doi

  15. [23]

    Model generalization in deep learning applications for land cover mapping,

    L. Hu, C. Robinson, and B. Dilkina, “Model generalization in deep learning applications for land cover mapping,” arXiv preprint arXiv:2008.10351, 2020

    2008 arXiv

  16. [24]

    Bringing back the context: Camera trap species identification as link prediction on multimodal knowledge graphs,

    V. Pahuja et al. , “Bringing back the context: Camera trap species identification as link prediction on multimodal knowledge graphs,” arXiv preprint arXiv:2401.00608, 2023

    2023 arXiv

  17. [25]

    Camera trapping —Advancing the technology,

    W.-M. Wong and S. Kachel, “Camera trapping —Advancing the technology,” in Snow Leopards, Elsevier, 2024, pp. 415–428. doi: 10.1016/B978-0-323-85775-8.00018-2

    2024 doi

  18. [26]

    Harnessing Artificial Intelligence for Wildlife Conservation,

    P. Fergus, C. Chalmers, S. Longmore, and S. Wich, “Harnessing Artificial Intelligence for Wildlife Conservation,” arXiv preprint arXiv:2409.10523, 2024. 20

    2024 arXiv

  19. [27]

    Yolov10: Real-time end-to-end object detection,

    A. Wang et al., “Yolov10: Real-time end-to-end object detection,” arXiv preprint arXiv:2405.14458, 2024

    2024 arXiv

  20. [28]

    Large scale fine-grained categorization and domain- specific transfer learning,

    Y. Cui, Y. Song, C. Sun, A. Howard, and S. Belongie, “Large scale fine-grained categorization and domain- specific transfer learning,” in Proceedings of the IEEE conference on computer vision and pattern recognition , 2018, pp. 4109–4118

    2018

  21. [29]

    Targeted transfer learning to improve performance in small medical physics datasets,

    M. Romero, Y. Interian, T. Solberg, and G. Valdes, “Targeted transfer learning to improve performance in small medical physics datasets,” Med Phys, vol. 47, no. 12, pp. 6246–6256, 2020

    2020

  22. [30]

    Optimizing High-Throughput Inference on Graph Neural Networks at Shared Computing Facilities with the NVIDIA Triton Inference Server,

    C. Savard et al., “Optimizing High-Throughput Inference on Graph Neural Networks at Shared Computing Facilities with the NVIDIA Triton Inference Server,” Comput Softw Big Sci, vol. 8, no. 1, p. 14, Dec. 2024, doi: 10.1007/s41781-024-00123-2

    2024 doi

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.