Camera Trapping Technology and Related Advances: into the New Millennium

Meek, Paul D.; Ballard, Guy; Falzon, Greg; Williamson, Jaimen; Milne, Heath; Farrell, Robert E.; Stover, Joshua D.; Mather-Zardain, Atalya T.; Bishop, James C.; Cheung, Elrond Ka-Wai; Lawson, Christopher; Munezero, Amos M.; Schneider, Derek; Johnston, Beau E.; Kiani, Ehsan; Shahinfar, Saleh; Sadgrove, Edmund J.; Fleming, Peter J. S.

doi:10.7882/az.2019.035

Cited by 13 publications

(17 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The rise of camera trapping as a scientific and management data collection tool has forged an unstoppable path across the globe (Meek et al, 2015;O'Connell, Nichols & Karanth, 2011;Rovero et al, 2013;Swann & Perkins, 2014). The consequence has created a hiatus between data capture and data storage/analysis because of the large volumes of image data collected and the manual analysis of that data (Falzon et al, 2014;Meek et al, 2019;Price-Tack et al2016;Scotson et al, 2017). Automating the analysis of camera trap data has been recognised as a critical step forward (Falzon et al, 2020;Meek et al, 2015;Nazir et al, 2017) although there are enormous difficulties, challenges and technological impediments to creating robust automated analytical tools (Falzon et al, 2020;Meek et al, 2019).…”

Section: Introductionmentioning

confidence: 99%

“…The consequence has created a hiatus between data capture and data storage/analysis because of the large volumes of image data collected and the manual analysis of that data (Falzon et al, 2014;Meek et al, 2019;Price-Tack et al2016;Scotson et al, 2017). Automating the analysis of camera trap data has been recognised as a critical step forward (Falzon et al, 2020;Meek et al, 2015;Nazir et al, 2017) although there are enormous difficulties, challenges and technological impediments to creating robust automated analytical tools (Falzon et al, 2020;Meek et al, 2019). This pressing need to develop reliable automated software to automatically scan, detect and sort data for further analysis has led to the age of big data and the subsequent integration of field ecology and ecological informatics.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

“How many images do I need?” Understanding how sample size per class affects deep learning model performance metrics for balanced designs in autonomous wildlife monitoring

Shahinfar

Meek

Falzon

2020

Ecological Informatics

Self Cite

102

View full text Add to dashboard Cite

1. Deep learning (DL) algorithms are the state of the art in automated classification of wildlife camera trap images. The challenge is that the ecologist cannot know in advance how many images per species they need to collect for model training in order to achieve their desired classification accuracy. In fact there is limited empirical evidence in the context of camera trapping to demonstrate that increasing sample size will lead to improved accuracy. 2. In this study we explore in depth the issues of deep learning model performance for progressively increasing per class (species) sample sizes. We also provide ecologists with an approximation formula to estimate how many images per animal species they need for certain accuracy level a priori. This will help ecologists for optimal allocation of resources, work and efficient study design. 3. In order to investigate the effect of number of training images; seven training sets with 10, 20, 50, 150, 500, 1000 images per class were designed. Six deep learning architectures namely ResNet-18, ResNet-50, ResNet-152, DnsNet-121, DnsNet-161, and DnsNet-201 were trained and tested on a common exclusive testing set of 250 images per class. The whole experiment was repeated on three similar datasets from Australia, Africa and North America and the results were compared. Simple regression equations for use by practitioners to approximate model performance metrics are provided. Generalizes additive models (GAM) are shown to be effective in modelling DL performance metrics based on the number of training images per class, tuning scheme and dataset. 4. Overall, our trained models classified images with 0.94 accuracy (ACC), 0.73 precision (PRC), 0.72 true positive rate (TPR), and 0.03 false positive rate (FPR). Variation in model performance metrics among datasets, species and deep learning architectures exist and are shown distinctively in the discussion section. The ordinary least squares regression models explained 57%, 54%, 52%, and 34% of expected variation of ACC, PRC, TPR, and FPR according to number of images available for training. Generalized additive models explained 77%, 69%, 70%, and 53% of deviance for ACC, PRC, TPR, and FPR respectively. 5. Predictive models were developed linking number of training images per class, model, dataset and performance metrics. The ordinary least squares regression and Generalised additive models developed provides a practical toolbox to estimate model performance with respect to different numbers of training images.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

“How many images do I need?” Understanding how sample size per class affects deep learning model performance metrics for balanced designs in autonomous wildlife monitoring

Shahinfar

Meek

Falzon

2020

Ecological Informatics

Self Cite

102

View full text Add to dashboard Cite

show abstract

“…There are various kinds of camera traps, and a detailed review of these can be found elsewhere (e.g. (Burton et al., 2015; Meek et al., 2019; Rovero, Zimmermann, Berzi, & Meek, 2013). The onset of digital photography has revolutionized camera trapping by allowing cameras to operate stealthily and for several days without having to replace film rolls.…”

Section: Introductionmentioning

confidence: 99%

Conservation and people: Towards an ethical code of conduct for the use of camera traps in wildlife research

Sharma

Fiechter²,

George³

et al. 2020

Ecol Sol and Evidence

View full text Add to dashboard Cite

1. Camera trapping is a widely employed tool in wildlife research, used to estimate animal abundances, understand animal movement, assess species richness and understand animal behaviour. In addition to images of wild animals, research cameras often record human images, inadvertently capturing behaviours ranging from innocuous actions to potentially serious crimes. 2. With the increasing use of camera traps, there is an urgent need to reflect on how researchers should deal with human images caught on cameras. On the one hand, it is important to respect the privacy of individuals caught on cameras, while, on the other hand, there is a larger public duty to report illegal activity. This creates ethical dilemmas for researchers. 3. Here, based on our camera-trap research on snow leopards Panthera uncia, we outline a general code of conduct to help improve the practice of camera trap based research and help researchers better navigate the ethical-legal tightrope of this important research tool. K E Y W O R D S camera trap, code of conduct, ethics, human rights, law, PARTNERS principles for communitybased conservation, privacy, snow leopard 1 INTRODUCTION Camera traps have become important tools for researchers, conservationists, and wildlife managers and are being used to study wildlife and urban ecology (Anton, Hartley, Geldenhuis, & Wittmer, 2018; O'Connell, Nichols, & Karanth, 2011). Camera trapping is especially valuable when research or management involves elusive species living in difficult to access habitats (O'Connell et al., 2011). Camera traps are used to estimate animal abundances, understand animal movement (Borchers, Distiller, Foster, Harmsen, & Milazzo, 2014), assess species This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

show abstract

“…Reducing the processing time of camera trap image analysis is fast becoming one of the most important drivers in the integration of computer science technologies and ecological research (Meek, Fleming et al 2014, Fegraus and MacCarthy 2016, Meek, Ballard et al 2019). Due to the increasing adoption of camera traps as a core survey method, the quantity of data collected by ecological practitioners has increased beyond that of any historical fauna survey method.…”

Section: Introductionmentioning

confidence: 99%

Location Invariant Animal Recognition Using Mixed Source Datasets and Deep Learning

Shepley

Falzon²,

Meek³

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

1.A time-consuming challenge faced by camera trap practitioners all over the world is the extraction of meaningful data from images to inform ecological management. The primary methods of image processing used by practitioners includes manual analysis and citizen science. An increasingly popular alternative is automated image classification software. However, most automated solutions are not sufficiently robust to be deployed on a large scale. Key challenges include limited access to images for each species and lack of location invariance when transferring models between sites. This prevents optimal use of ecological data and results in significant expenditure of time and resources to annotate and retrain deep learning models.2. In this study, we aimed to (a) assess the value of publicly available non-iconic FlickR images in the training of deep learning models for camera trap object detection, (b) develop an out-ofthe-box location invariant automated camera trap image processing solution for ecologist using deep transfer learning and (c) explore the use of small subsets of camera trap images in optimisation of a FlickR trained deep learning model for high precision ecological object detection.3. We collected and annotated a dataset of images of "pigs" (Sus scrofa and Phacochoerus africanus) from the consumer image sharing website FlickR. These images were used to achieve transfer learning using a RetinaNet model in the task of object detection. We compared the performance of this model to the performance of models trained on combinations of camera trap images obtained from five different projects, each characterised by 5 different geographical regions. Furthermore, we explored optimisation of the FlickR model via infusion of small subsets of camera trap images to increase robustness in difficult images. 4. In most cases, the mean Average Precision (mAP) of the FlickR trained model when tested on out of sample camera trap sites (67.21-91.92%) was significantly higher than the mAP achieved by models trained on only one geographical location (4.42-90.8%) and rivalled the mAP of models trained on mixed camera trap datasets (68.96-92.75%). The infusion of camera trap images into the FlickR training further improved AP by 5.10-22.32% to 83.60-97.02%. Ecology researchers can use FlickR images in the training of automated deep learningsolutions for camera trap image processing to significantly reduce time and resource expenditure by allowing the development of location invariant, highly robust out-of-the-box solutions. This would allow AI technologies to be deployed on a large scale in ecological applications.

show abstract

Camera Trapping Technology and Related Advances: into the New Millennium

Cited by 13 publications

References 0 publications

“How many images do I need?” Understanding how sample size per class affects deep learning model performance metrics for balanced designs in autonomous wildlife monitoring

“How many images do I need?” Understanding how sample size per class affects deep learning model performance metrics for balanced designs in autonomous wildlife monitoring

Conservation and people: Towards an ethical code of conduct for the use of camera traps in wildlife research

Location Invariant Animal Recognition Using Mixed Source Datasets and Deep Learning

Contact Info

Product

Resources

About