Computer Vision for Wildlife Monitoring: Detecting Brown Howler Monkeys using YOLO

Gabriel Ferri Schneider; Guido Luis Glufke Mainardi; J\'ulio C\'esar Bicca-Marques; M\'arcia Jardim; Patr\'icia Dias; Paulo Ricardo Knob; Soraia Raupp Musse

arxiv: 2607.01396 · v1 · pith:KFLUSFYQnew · submitted 2026-07-01 · 💻 cs.CV · cs.GR

Computer Vision for Wildlife Monitoring: Detecting Brown Howler Monkeys using YOLO

Gabriel Ferri Schneider , Guido Luis Glufke Mainardi , Paulo Ricardo Knob , Patr\'icia Dias , M\'arcia Jardim , J\'ulio C\'esar Bicca-Marques , Soraia Raupp Musse This is my paper

Pith reviewed 2026-07-03 20:57 UTC · model grok-4.3

classification 💻 cs.CV cs.GR

keywords brown howler monkeysYOLOv10camera trapsobject detectionwildlife monitoringcanopy bridgesauxiliary dataconservation

0 comments

The pith

Fine-tuning YOLOv10 with auxiliary data improves brown howler monkey detection in camera trap videos.

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

The paper tests whether adding auxiliary images from other sources can improve a YOLOv10 model trained to spot brown howler monkeys in camera trap footage from canopy bridges. Urban expansion fragments forests and increases risks for arboreal animals, so bridges help but need monitoring. Camera traps generate many false positives that take time to review manually. By trying different mixes of auxiliary data during fine-tuning, the work shows how computer vision can automate species detection to support conservation. This matters because effective monitoring can confirm if bridges reduce mortality from habitat fragmentation.

Core claim

The authors fine-tune the YOLOv10 object detection framework on brown howler monkey images from camera traps, incorporating varying proportions of auxiliary data to enhance model performance for identifying the species in videos.

What carries the argument

YOLOv10 framework fine-tuned with varying proportions of auxiliary data for object detection in camera trap videos.

Load-bearing premise

Auxiliary data from other sources will enhance performance on the specific camera-trap videos without causing domain shift or bias that harms accuracy.

What would settle it

A test showing that models trained with auxiliary data have lower precision or recall on held-out camera trap images compared to those trained only on target data would falsify the improvement claim.

Figures

Figures reproduced from arXiv: 2607.01396 by Gabriel Ferri Schneider, Guido Luis Glufke Mainardi, J\'ulio C\'esar Bicca-Marques, M\'arcia Jardim, Patr\'icia Dias, Paulo Ricardo Knob, Soraia Raupp Musse.

**Figure 1.** Figure 1: Image examples from the Primary Dataset. 3.2 Auxiliary Datasets In order to explore the use of diversity in the network training, besides our primary dataset, we obtained two additional public datasets from Roboflow Universe2 . The first is the “Human Detection Dataset”, which is comprised of 5,000 images containing people, annotated for object detection. The second is the “Non-Human Primate Dataset” and … view at source ↗

**Figure 2.** Figure 2: Image examples from the Auxiliary Datasets. On the top row, two examples from “Human Detection Dataset”. On the bottom row, two examples from “Non-Human Primate Dataset”. datasets, we developed a synthetic dataset using Unity3 to increase environmental diversity and test the impact of synthetic data on fine-tuning. The development process was as follows: • 3D model development: A fully textured 3D brown h… view at source ↗

**Figure 4.** Figure 4: Scenario examples created for the synthetic dataset generation. 3.3 Video Triage In order to evaluate the model’s ability to filter out videos that do not contain brown howler monkeys or other useful information, without discarding those in which the animals appear, we designed a procedure to assess the model’s ability to triage videos. Starting from the 16,179 camera-trap videos (Section 3.1), we system… view at source ↗

**Figure 5.** Figure 5: Detection performance of the three variations of auxiliary data, namely Non-Human Primates, Human, and Synthetic Data, for F1-Score. As commented in Section 4.1, we also per- [PITH_FULL_IMAGE:figures/full_fig_p007_5.png] view at source ↗

**Figure 6.** Figure 6: Detection performance of the three variations of auxiliary data, namely Non-Human Primates, Human, and Synthetic Data, for mAP@0.5. formed an experiment with different combinations of synthetic and real images. The results of this additional fine-tunings, using synthetic data, are presented in [PITH_FULL_IMAGE:figures/full_fig_p008_6.png] view at source ↗

read the original abstract

Urban expansion threatens global biodiversity, especially affecting arboreal species due to the fragmentation of forest habitats. The movement of arboreal species across disjointed forest patches increases mortality risk and, thus, compromises their conservation. In this context, the installation of canopy bridges can be a viable strategy; yet continuous monitoring of their use by arboreal species is essential for ensuring their effectiveness, typically carried out with the aid of camera traps. However, this method often produces false-positive images that demand time from conservationists for review. In this context, computer vision algorithms can optimize the task of detecting target species using the canopy bridges. In this study, we explored the automatic detection of brown howler monkeys (Alouatta guariba) in videos obtained by camera traps. Given the need for a large number of annotated images of the target animals to train the algorithms, we tested the incorporation of auxiliary data to improve detection models, fine-tuning the YOLOv10 framework using varying proportions of them. The improvement of these automatic detection techniques contributes to conservation efforts, by providing automatic tools to monitor solutions that minimize the impact of human interference in animals habitats.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

Standard YOLOv10 fine-tuning on brown howler monkey camera-trap videos with auxiliary data tests, but abstract supplies zero metrics so the improvement claim cannot be checked.

read the letter

The paper applies YOLOv10 to automatic detection of brown howler monkeys in canopy-bridge camera-trap videos and tests whether mixing in different proportions of auxiliary data during fine-tuning helps on the target domain.

This is a direct, practical response to a real monitoring bottleneck: camera traps produce many false positives that eat up reviewer time, and canopy bridges are one tool for reducing habitat fragmentation effects. Using auxiliary data when labeled target images are scarce is a common tactic in applied detection work, and testing varying proportions is a reasonable way to probe how much extra data is useful.

The clear weakness is that the abstract states the method and the auxiliary-data experiment but reports no accuracy figures, no baseline comparisons, no dataset sizes, no error bars, and no validation protocol. Without those numbers the central claim about improvement cannot be evaluated, and the risk of domain shift from the auxiliary data stays untested. The full manuscript may contain the results, but nothing in the provided text lets a reader judge soundness.

The work is an empirical extension of an existing detector to one new species and context rather than a methodological advance. It will interest conservation biologists or applied CV groups who need off-the-shelf tools for similar camera-trap tasks. A reader already familiar with YOLO transfer learning will not learn new techniques here.

If the full paper supplies clear metrics, reproducible splits, and honest baselines, it is worth sending to referees; otherwise the lack of evidence makes it a routine application note rather than something that needs detailed review.

Referee Report

1 major / 2 minor

Summary. The manuscript describes an application of the YOLOv10 object detection framework to automatically detect brown howler monkeys (Alouatta guariba) in camera-trap videos collected at canopy bridges. The central experiment tests whether fine-tuning with varying proportions of auxiliary data improves detection performance relative to the target domain alone, with the goal of reducing false-positive images that require manual review by conservationists.

Significance. If the empirical results demonstrate reliable gains from auxiliary data without harmful domain shift, the work would supply a practical, deployable tool for scaling camera-trap monitoring of arboreal species. The approach directly addresses a documented bottleneck in conservation workflows. However, the absence of any quantitative results, baselines, or validation protocol in the supplied text prevents evaluation of whether this potential is realized.

major comments (1)

[Abstract] Abstract: the central claim that 'incorporation of auxiliary data' improves detection models is stated, yet the abstract (and the provided manuscript text) supplies no performance metrics, dataset sizes, train/test splits, evaluation protocol, baselines, or error bars. Without these, the claim cannot be assessed and the paper's contribution remains unverifiable.

minor comments (2)

The manuscript would benefit from a dedicated Methods section that specifies the source and annotation protocol for the auxiliary data, the exact YOLOv10 variant and training hyperparameters, and the definition of 'varying proportions'.
Figure and table captions should be expanded to include the precise metrics plotted and the number of runs or cross-validation folds used.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their review and for identifying the need for quantitative details to support the central claims. We agree that the abstract and manuscript as supplied do not contain performance metrics, dataset information, or evaluation details, which prevents verification of the contribution. We will revise the manuscript accordingly.

read point-by-point responses

Referee: [Abstract] Abstract: the central claim that 'incorporation of auxiliary data' improves detection models is stated, yet the abstract (and the provided manuscript text) supplies no performance metrics, dataset sizes, train/test splits, evaluation protocol, baselines, or error bars. Without these, the claim cannot be assessed and the paper's contribution remains unverifiable.

Authors: We agree that the supplied abstract and manuscript text lack the required quantitative elements. The experiments in the work compare YOLOv10 fine-tuned on target camera-trap data alone versus with varying proportions of auxiliary data from related domains, but these results (including mAP, precision, recall, and false-positive reduction) are not reported in the abstract or described with protocol details in the provided text. We will revise the abstract to include key metrics (e.g., the mAP improvement and false-positive reduction achieved with auxiliary data) and expand the methods/results sections to specify dataset sizes, train/test splits, the evaluation protocol (target-domain hold-out testing), baselines (YOLOv10 trained only on target data), and any error bars or repeated-run statistics. This will make the claims verifiable. revision: yes

Circularity Check

0 steps flagged

No significant circularity

full rationale

The paper is a purely empirical application of the existing YOLOv10 model to camera-trap video for brown howler monkey detection. It tests the effect of adding varying proportions of auxiliary data during fine-tuning but contains no equations, derivations, parameter fits presented as predictions, uniqueness theorems, or self-citations that bear load on any central claim. The abstract and described content supply only experimental methodology and motivation; no step reduces by construction to its own inputs. This is the expected outcome for an applied CV paper without theoretical content.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Only the abstract is available; the paper relies on standard machine-learning assumptions for fine-tuning object detectors but introduces no explicit free parameters, new axioms, or invented entities.

axioms (1)

domain assumption YOLOv10 can be fine-tuned on camera-trap imagery for species detection
Standard assumption invoked by the choice to fine-tune the framework.

pith-pipeline@v0.9.1-grok · 5764 in / 1031 out tokens · 24510 ms · 2026-07-03T20:57:31.959149+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

27 extracted references · 27 canonical work pages

[1]

Wildlife is imperiled in peri-urban landscapes: threats to arboreal mammals.Science of the Total Environment, 821:152883, 2022

´Oscar M Chaves, JC Souza J ´unior, Ger- son Buss, Zelinda MB Hirano, MMA Jardim, ELS Amaral, Jo˜ao Cl´audio Godoy, Amanda Rezende Peruchi, Thais Michel, and J´ulio C´esar Bicca-Marques. Wildlife is imperiled in peri-urban landscapes: threats to arboreal mammals.Science of the Total Environment, 821:152883, 2022

work page 2022
[2]

Ecological impact and cost-effectiveness of wildlife crossings in a highly fragmented landscape: a multi- method approach.Landscape Ecology, 35(7):1701–1720, 2020

Frans J Sijtsma, Eelke van der Veen, Ar- jen van Hinsberg, Rogier Pouwels, Ren ´ee Bekker, Ren ´e E van Dijk, Mark Grut- ters, Raymond Klaassen, Margriet Krijn, Maarten Mouissie, et al. Ecological impact and cost-effectiveness of wildlife crossings in a highly fragmented landscape: a multi- method approach.Landscape Ecology, 35(7):1701–1720, 2020

work page 2020
[3]

Fredrick Lala, Patrick I Chiyo, Patrick Omondi, Benson Okita-Ouma, Erustus Kanga, Michael Koskei, Lydia Tiller, Aaron W Morris, William J Severud, and Joseph K Bump. Influence of infrastruc- ture, ecology, and underpass-dimensions on multi-year use of standard gauge rail- way underpasses by mammals in tsavo, kenya.Scientific reports, 12(1):5698, 2022

work page 2022
[4]

Do wildlife crossings mitigate the roadkill mortality of tropical mammals? a case study from costa rica.Diversity, 14(8):665, 2022

Ronald Villalobos-Hoffman, Jack E Ew- ing, and Michael S Mooring. Do wildlife crossings mitigate the roadkill mortality of tropical mammals? a case study from costa rica.Diversity, 14(8):665, 2022

work page 2022
[5]

Methods to establish canopy bridges to increase natural connec- tivity in linear infrastructure development

Tremaine Gregory, Farah Carrasco-Rueda, Jessica Deichmann, Joseph Kolowski, Marcel Costa Faura, Francisco Dallmeier, and Alfonso Alonso. Methods to establish canopy bridges to increase natural connec- tivity in linear infrastructure development. Society of Petroleum Engineers Journal, page 1–10, jun 2013

work page 2013
[6]

ah lai’s crossing

Jo Leen Yap, Nik Fadzly Nik Rosely, Mazrul Mahadzir, Mark Louis Benedict, Vikneswaran Muniandy, and Nadine Rup- pert. “ah lai’s crossing” – malaysia’s first artificial road canopy bridge to facilitate safer arboreal wildlife crossings.Folia Pri- matologica, 93(3-6):255–269, mar 2022

work page 2022
[7]

Tremaine Gregory, Fernanda Abra, Birthe Linden, K. A. I. Nekaris, Kylie Soanes, and Fernanda Z. Teixeira. A new window into canopy bridges as a mitigation strat- egy for arboreal mammals.Folia Primato- logica, 93(3-6):197–203, dez 2022

work page 2022
[8]

Arbo- real wildlife bridges in the tropical rainfor- est of costa rica’s osa peninsula.Folia Pri- matologica, 93:1–17, abr 2022

Eleanor Flatt, Arianna Basto, Carolina Pinto, Johan Ortiz, Kassandra Navarro, Neil Reed, Hilary Brumberg, Marco Chaverri, and Andrew Whitworth. Arbo- real wildlife bridges in the tropical rainfor- est of costa rica’s osa peninsula.Folia Pri- matologica, 93:1–17, abr 2022

work page 2022
[9]

Canopy bridges as road overpasses for wildlife in urban fragmented landscapes.Biota Neotrop., 13(1):117–123, mar 2013

Fernanda Zimmermann Teixeira, Rodrigo Cambar ´a Printes, Jo ˜ao Cl´audio Godoy Fagundes, Andr ´e Chein Alonso, and Andreas Kindel. Canopy bridges as road overpasses for wildlife in urban fragmented landscapes.Biota Neotrop., 13(1):117–123, mar 2013

work page 2013
[10]

Ungulate use of an at-grade fauna passage and roadside animal detection sys- tem: A pilot study from southern sweden

Manisha Bhardwaj, Filippa Erixon, Isak Holmberg, Andreas Seiler, Emma H˚akansson, Marcus Elfstr ¨om, and Mattias Olsson. Ungulate use of an at-grade fauna passage and roadside animal detection sys- tem: A pilot study from southern sweden. Front. Environ. Sci., 10(991551):991551, dez 2022

work page 2022
[11]

Bridging the gap: assessing the effective- ness of rope bridges for wildlife in singa- pore.Folia Primatologica, 93:1–12, abr 2022

Sebastian Ow, Sharon Chan, Yuet Toh, Su Chan, Jayasri Lakshminarayanan, Sab- rina Jabbar, Andie Ang, and Adrian Loo. Bridging the gap: assessing the effective- ness of rope bridges for wildlife in singa- pore.Folia Primatologica, 93:1–12, abr 2022

work page 2022
[12]

Deep learning for automatic facial detection and recog- nition in japanese macaques: illuminating social networks.Primates, 65(4):265–279, 2024

Julien Paulet, Axel Molina, Benjamin Beltzung, Takafumi Suzumura, Shinya Ya- mamoto, and C´edric Sueur. Deep learning for automatic facial detection and recog- nition in japanese macaques: illuminating social networks.Primates, 65(4):265–279, 2024

work page 2024
[13]

Primate face identifica- tion in the wild

Ankita Shukla, Gullal Singh Cheema, Saket Anand, Qamar Qureshi, and Yad- vendradev Jhala. Primate face identifica- tion in the wild. InPacific Rim Inter- national Conference on Artificial Intelli- gence, pages 387–401. Springer, 2019

work page 2019
[14]

Towards automatic detection of monkey faces

Manning Zhang, Susu Guo, and Xiao- hua Xie. Towards automatic detection of monkey faces. In24th International Conference on Pattern Recognition, page 2564–2569, 2018

work page 2018
[15]

A survey on data collection for machine learning: a big data-ai integration perspective.IEEE Transactions on Knowl- edge and Data Engineering, 33(4):1328– 1347, 2019

Yuji Roh, Geon Heo, and Steven Euijong Whang. A survey on data collection for machine learning: a big data-ai integration perspective.IEEE Transactions on Knowl- edge and Data Engineering, 33(4):1328– 1347, 2019

work page 2019
[16]

A survey on deep learn- ing tools dealing with data scarcity: defi- nitions, challenges, solutions, tips, and ap- plications.Journal of Big Data, 10(1):46, 2023

Laith Alzubaidi, Jinshuai Bai, Aiman Al- Sabaawi, Jose Santamar ´ıa, Ahmed Shi- hab Albahri, Bashar Sami Nayyef Al- Dabbagh, Mohammed A Fadhel, Mo- hamed Manoufali, Jinglan Zhang, Ali H Al-Timemy, et al. A survey on deep learn- ing tools dealing with data scarcity: defi- nitions, challenges, solutions, tips, and ap- plications.Journal of Big Data, 10(1):46, 2023

work page 2023
[17]

Yolov10: Real-time end-to-end object detection.Ad- vances in Neural Information Processing Systems, 37:107984–108011, 2024

Ao Wang, Hui Chen, Lihao Liu, Kai Chen, Zijia Lin, Jungong Han, et al. Yolov10: Real-time end-to-end object detection.Ad- vances in Neural Information Processing Systems, 37:107984–108011, 2024

work page 2024
[18]

Computer vision for primate behavior ana- lysis in the wild.Nature Methods, pages 1–13, 2025

Richard V ogg, Timo L ¨uddecke, Jonathan Henrich, Sharmita Dey, Matthias Nuske, Valentin Hassler, Derek Murphy, Julia Fis- cher, Julia Ostner, Oliver Sch ¨ulke, et al. Computer vision for primate behavior ana- lysis in the wild.Nature Methods, pages 1–13, 2025

work page 2025
[19]

Indoor synthetic data generation: A systematic review.Com- puter Vision and Image Understanding, 240:103907, 2024

Hannah Schieber, Kubilay Can Demir, Constantin Kleinbeck, Seung Hee Yang, and Daniel Roth. Indoor synthetic data generation: A systematic review.Com- puter Vision and Image Understanding, 240:103907, 2024

work page 2024
[20]

Motsynth: How can synthetic data help pedestrian detection and tracking? InPro- ceedings of the IEEE/CVF International Conference on Computer Vision, pages 10849–10859, 2021

Matteo Fabbri, Guillem Bras ´o, Gianluca Maugeri, Orcun Cetintas, Riccardo Gas- parini, Aljo ˇsa O ˇsep, Simone Calderara, Laura Leal-Taix ´e, and Rita Cucchiara. Motsynth: How can synthetic data help pedestrian detection and tracking? InPro- ceedings of the IEEE/CVF International Conference on Computer Vision, pages 10849–10859, 2021

work page 2021
[21]

Ex- ploring the use of synthetic data to improve pedestrian detection in cnn

Rubens Montanha and Soraia Musse. Ex- ploring the use of synthetic data to improve pedestrian detection in cnn. In2025 38th SIBGRAPI Conference on Graphics, Pat- terns and Images (SIBGRAPI), pages 1–6. IEEE, 2025

work page 2025
[22]

Face detection of golden mon- keys via regional color quantization and incremental self-paced curriculum learn- ing.Multimedia Tools and Applications, 77(3):3143–3170, 2018

Pengfei Xu, Songtao Guo, Qiguang Miao, Baoguo Li, Xiaojiang Chen, and Dingyi Fang. Face detection of golden mon- keys via regional color quantization and incremental self-paced curriculum learn- ing.Multimedia Tools and Applications, 77(3):3143–3170, 2018

work page 2018
[23]

Evaluation of the effect of transfer learning to multi- instance detection of monkeys

Riza Rae Pineda, Takatomi Kubo, Masaki Shimada, and Kazushi Ikeda. Evaluation of the effect of transfer learning to multi- instance detection of monkeys. In2021 Asia-Pacific Signal and Information Pro- cessing Association Annual Summit and Conference (APSIPA ASC), pages 1357–

work page
[24]

Behavior and safety in the use of canopy bridges by howler monkeys in an experimental environment.Applied Animal Behaviour Science, 290:106696, 2025

Patr ´ıcia Dias, Fernanda Zimmermann Teixeira, Victor Hugo Campos Ferrugem, Itatiele Farias Vivian, and M ´arcia Maria de Assis Jardim. Behavior and safety in the use of canopy bridges by howler monkeys in an experimental environment.Applied Animal Behaviour Science, 290:106696, 2025

work page 2025
[25]

The via annotation software for images, audio and video

Abhishek Dutta and Andrew Zisserman. The via annotation software for images, audio and video. InProceedings of the 27th ACM International Conference on Multimedia, page 2276–2279, 2019

work page 2019
[26]

Monster mash: a single-view approach to casual 3d modeling and animation.ACM Trans- actions on Graphics (ToG), 39(6):1–12, 2020

Marek Dvoro ˇzˇn´ak, Daniel S `ykora, Cas- sidy Curtis, Brian Curless, Olga Sorkine- Hornung, and David Salesin. Monster mash: a single-view approach to casual 3d modeling and animation.ACM Trans- actions on Graphics (ToG), 39(6):1–12, 2020

work page 2020
[27]

Evaluating time series forecast- ing models: An empirical study on per- formance estimation methods.Machine Learning, 109(11):1997–2028, 2020

Vitor Cerqueira, Luis Torgo, and Igor Mozetiˇc. Evaluating time series forecast- ing models: An empirical study on per- formance estimation methods.Machine Learning, 109(11):1997–2028, 2020

work page 1997

[1] [1]

Wildlife is imperiled in peri-urban landscapes: threats to arboreal mammals.Science of the Total Environment, 821:152883, 2022

´Oscar M Chaves, JC Souza J ´unior, Ger- son Buss, Zelinda MB Hirano, MMA Jardim, ELS Amaral, Jo˜ao Cl´audio Godoy, Amanda Rezende Peruchi, Thais Michel, and J´ulio C´esar Bicca-Marques. Wildlife is imperiled in peri-urban landscapes: threats to arboreal mammals.Science of the Total Environment, 821:152883, 2022

work page 2022

[2] [2]

Ecological impact and cost-effectiveness of wildlife crossings in a highly fragmented landscape: a multi- method approach.Landscape Ecology, 35(7):1701–1720, 2020

Frans J Sijtsma, Eelke van der Veen, Ar- jen van Hinsberg, Rogier Pouwels, Ren ´ee Bekker, Ren ´e E van Dijk, Mark Grut- ters, Raymond Klaassen, Margriet Krijn, Maarten Mouissie, et al. Ecological impact and cost-effectiveness of wildlife crossings in a highly fragmented landscape: a multi- method approach.Landscape Ecology, 35(7):1701–1720, 2020

work page 2020

[3] [3]

Fredrick Lala, Patrick I Chiyo, Patrick Omondi, Benson Okita-Ouma, Erustus Kanga, Michael Koskei, Lydia Tiller, Aaron W Morris, William J Severud, and Joseph K Bump. Influence of infrastruc- ture, ecology, and underpass-dimensions on multi-year use of standard gauge rail- way underpasses by mammals in tsavo, kenya.Scientific reports, 12(1):5698, 2022

work page 2022

[4] [4]

Do wildlife crossings mitigate the roadkill mortality of tropical mammals? a case study from costa rica.Diversity, 14(8):665, 2022

Ronald Villalobos-Hoffman, Jack E Ew- ing, and Michael S Mooring. Do wildlife crossings mitigate the roadkill mortality of tropical mammals? a case study from costa rica.Diversity, 14(8):665, 2022

work page 2022

[5] [5]

Methods to establish canopy bridges to increase natural connec- tivity in linear infrastructure development

Tremaine Gregory, Farah Carrasco-Rueda, Jessica Deichmann, Joseph Kolowski, Marcel Costa Faura, Francisco Dallmeier, and Alfonso Alonso. Methods to establish canopy bridges to increase natural connec- tivity in linear infrastructure development. Society of Petroleum Engineers Journal, page 1–10, jun 2013

work page 2013

[6] [6]

ah lai’s crossing

Jo Leen Yap, Nik Fadzly Nik Rosely, Mazrul Mahadzir, Mark Louis Benedict, Vikneswaran Muniandy, and Nadine Rup- pert. “ah lai’s crossing” – malaysia’s first artificial road canopy bridge to facilitate safer arboreal wildlife crossings.Folia Pri- matologica, 93(3-6):255–269, mar 2022

work page 2022

[7] [7]

Tremaine Gregory, Fernanda Abra, Birthe Linden, K. A. I. Nekaris, Kylie Soanes, and Fernanda Z. Teixeira. A new window into canopy bridges as a mitigation strat- egy for arboreal mammals.Folia Primato- logica, 93(3-6):197–203, dez 2022

work page 2022

[8] [8]

Arbo- real wildlife bridges in the tropical rainfor- est of costa rica’s osa peninsula.Folia Pri- matologica, 93:1–17, abr 2022

Eleanor Flatt, Arianna Basto, Carolina Pinto, Johan Ortiz, Kassandra Navarro, Neil Reed, Hilary Brumberg, Marco Chaverri, and Andrew Whitworth. Arbo- real wildlife bridges in the tropical rainfor- est of costa rica’s osa peninsula.Folia Pri- matologica, 93:1–17, abr 2022

work page 2022

[9] [9]

Canopy bridges as road overpasses for wildlife in urban fragmented landscapes.Biota Neotrop., 13(1):117–123, mar 2013

Fernanda Zimmermann Teixeira, Rodrigo Cambar ´a Printes, Jo ˜ao Cl´audio Godoy Fagundes, Andr ´e Chein Alonso, and Andreas Kindel. Canopy bridges as road overpasses for wildlife in urban fragmented landscapes.Biota Neotrop., 13(1):117–123, mar 2013

work page 2013

[10] [10]

Ungulate use of an at-grade fauna passage and roadside animal detection sys- tem: A pilot study from southern sweden

Manisha Bhardwaj, Filippa Erixon, Isak Holmberg, Andreas Seiler, Emma H˚akansson, Marcus Elfstr ¨om, and Mattias Olsson. Ungulate use of an at-grade fauna passage and roadside animal detection sys- tem: A pilot study from southern sweden. Front. Environ. Sci., 10(991551):991551, dez 2022

work page 2022

[11] [11]

Bridging the gap: assessing the effective- ness of rope bridges for wildlife in singa- pore.Folia Primatologica, 93:1–12, abr 2022

Sebastian Ow, Sharon Chan, Yuet Toh, Su Chan, Jayasri Lakshminarayanan, Sab- rina Jabbar, Andie Ang, and Adrian Loo. Bridging the gap: assessing the effective- ness of rope bridges for wildlife in singa- pore.Folia Primatologica, 93:1–12, abr 2022

work page 2022

[12] [12]

Deep learning for automatic facial detection and recog- nition in japanese macaques: illuminating social networks.Primates, 65(4):265–279, 2024

Julien Paulet, Axel Molina, Benjamin Beltzung, Takafumi Suzumura, Shinya Ya- mamoto, and C´edric Sueur. Deep learning for automatic facial detection and recog- nition in japanese macaques: illuminating social networks.Primates, 65(4):265–279, 2024

work page 2024

[13] [13]

Primate face identifica- tion in the wild

Ankita Shukla, Gullal Singh Cheema, Saket Anand, Qamar Qureshi, and Yad- vendradev Jhala. Primate face identifica- tion in the wild. InPacific Rim Inter- national Conference on Artificial Intelli- gence, pages 387–401. Springer, 2019

work page 2019

[14] [14]

Towards automatic detection of monkey faces

Manning Zhang, Susu Guo, and Xiao- hua Xie. Towards automatic detection of monkey faces. In24th International Conference on Pattern Recognition, page 2564–2569, 2018

work page 2018

[15] [15]

A survey on data collection for machine learning: a big data-ai integration perspective.IEEE Transactions on Knowl- edge and Data Engineering, 33(4):1328– 1347, 2019

Yuji Roh, Geon Heo, and Steven Euijong Whang. A survey on data collection for machine learning: a big data-ai integration perspective.IEEE Transactions on Knowl- edge and Data Engineering, 33(4):1328– 1347, 2019

work page 2019

[16] [16]

A survey on deep learn- ing tools dealing with data scarcity: defi- nitions, challenges, solutions, tips, and ap- plications.Journal of Big Data, 10(1):46, 2023

Laith Alzubaidi, Jinshuai Bai, Aiman Al- Sabaawi, Jose Santamar ´ıa, Ahmed Shi- hab Albahri, Bashar Sami Nayyef Al- Dabbagh, Mohammed A Fadhel, Mo- hamed Manoufali, Jinglan Zhang, Ali H Al-Timemy, et al. A survey on deep learn- ing tools dealing with data scarcity: defi- nitions, challenges, solutions, tips, and ap- plications.Journal of Big Data, 10(1):46, 2023

work page 2023

[17] [17]

Yolov10: Real-time end-to-end object detection.Ad- vances in Neural Information Processing Systems, 37:107984–108011, 2024

Ao Wang, Hui Chen, Lihao Liu, Kai Chen, Zijia Lin, Jungong Han, et al. Yolov10: Real-time end-to-end object detection.Ad- vances in Neural Information Processing Systems, 37:107984–108011, 2024

work page 2024

[18] [18]

Computer vision for primate behavior ana- lysis in the wild.Nature Methods, pages 1–13, 2025

Richard V ogg, Timo L ¨uddecke, Jonathan Henrich, Sharmita Dey, Matthias Nuske, Valentin Hassler, Derek Murphy, Julia Fis- cher, Julia Ostner, Oliver Sch ¨ulke, et al. Computer vision for primate behavior ana- lysis in the wild.Nature Methods, pages 1–13, 2025

work page 2025

[19] [19]

Indoor synthetic data generation: A systematic review.Com- puter Vision and Image Understanding, 240:103907, 2024

Hannah Schieber, Kubilay Can Demir, Constantin Kleinbeck, Seung Hee Yang, and Daniel Roth. Indoor synthetic data generation: A systematic review.Com- puter Vision and Image Understanding, 240:103907, 2024

work page 2024

[20] [20]

Motsynth: How can synthetic data help pedestrian detection and tracking? InPro- ceedings of the IEEE/CVF International Conference on Computer Vision, pages 10849–10859, 2021

Matteo Fabbri, Guillem Bras ´o, Gianluca Maugeri, Orcun Cetintas, Riccardo Gas- parini, Aljo ˇsa O ˇsep, Simone Calderara, Laura Leal-Taix ´e, and Rita Cucchiara. Motsynth: How can synthetic data help pedestrian detection and tracking? InPro- ceedings of the IEEE/CVF International Conference on Computer Vision, pages 10849–10859, 2021

work page 2021

[21] [21]

Ex- ploring the use of synthetic data to improve pedestrian detection in cnn

Rubens Montanha and Soraia Musse. Ex- ploring the use of synthetic data to improve pedestrian detection in cnn. In2025 38th SIBGRAPI Conference on Graphics, Pat- terns and Images (SIBGRAPI), pages 1–6. IEEE, 2025

work page 2025

[22] [22]

Face detection of golden mon- keys via regional color quantization and incremental self-paced curriculum learn- ing.Multimedia Tools and Applications, 77(3):3143–3170, 2018

Pengfei Xu, Songtao Guo, Qiguang Miao, Baoguo Li, Xiaojiang Chen, and Dingyi Fang. Face detection of golden mon- keys via regional color quantization and incremental self-paced curriculum learn- ing.Multimedia Tools and Applications, 77(3):3143–3170, 2018

work page 2018

[23] [23]

Evaluation of the effect of transfer learning to multi- instance detection of monkeys

Riza Rae Pineda, Takatomi Kubo, Masaki Shimada, and Kazushi Ikeda. Evaluation of the effect of transfer learning to multi- instance detection of monkeys. In2021 Asia-Pacific Signal and Information Pro- cessing Association Annual Summit and Conference (APSIPA ASC), pages 1357–

work page

[24] [24]

Behavior and safety in the use of canopy bridges by howler monkeys in an experimental environment.Applied Animal Behaviour Science, 290:106696, 2025

Patr ´ıcia Dias, Fernanda Zimmermann Teixeira, Victor Hugo Campos Ferrugem, Itatiele Farias Vivian, and M ´arcia Maria de Assis Jardim. Behavior and safety in the use of canopy bridges by howler monkeys in an experimental environment.Applied Animal Behaviour Science, 290:106696, 2025

work page 2025

[25] [25]

The via annotation software for images, audio and video

Abhishek Dutta and Andrew Zisserman. The via annotation software for images, audio and video. InProceedings of the 27th ACM International Conference on Multimedia, page 2276–2279, 2019

work page 2019

[26] [26]

Monster mash: a single-view approach to casual 3d modeling and animation.ACM Trans- actions on Graphics (ToG), 39(6):1–12, 2020

Marek Dvoro ˇzˇn´ak, Daniel S `ykora, Cas- sidy Curtis, Brian Curless, Olga Sorkine- Hornung, and David Salesin. Monster mash: a single-view approach to casual 3d modeling and animation.ACM Trans- actions on Graphics (ToG), 39(6):1–12, 2020

work page 2020

[27] [27]

Evaluating time series forecast- ing models: An empirical study on per- formance estimation methods.Machine Learning, 109(11):1997–2028, 2020

Vitor Cerqueira, Luis Torgo, and Igor Mozetiˇc. Evaluating time series forecast- ing models: An empirical study on per- formance estimation methods.Machine Learning, 109(11):1997–2028, 2020

work page 1997