Wildlife conservation relies on sound knowledge of population information. This information is crucial for addressing questions related to community, ecosystem function, population dynamics, and behavioural ecology. It is obtained via studies that recognise individuals. A commonly used technique to achieve individual species recognition is the use of invasive methods that apply a tag to the animal's body. These methods have been applied to both marine and terrestrial species to assess both theoretical and applied questions [5, 6]. However, invasive procedures are expensive to implement and potentially reduce the animal's natural behaviour and performance, disturb its activities and relationship to others. In addition to being impracticable with large populations, invasive methods cause ethical and welfare con icts due to temporary or permanent application of tags. Alternatively, many species have body marks such as spots and ns that are individual-speci c and hence can be utilised for individual recognition. These methods are cost-e ective and not harmful to the animal's life. These approaches have evolved as a reliable alternative to invasive methods and have been applied to a range of animals including mammals, amphibians, reptiles and shes [8, 9]. This paper aims at developing a machine learning algorithm which exploits individual-speci c marks to automate the individual identi cation task and compares the model results with some of the existing computer-aided software used by the ecology community. The developed model is tested on two case studies, a humpback whale (HBW) dataset and a western leopard toad (WLT) dataset. The HBW dataset consists of 25 631 images from 14 668 individuals. They are originally collected by various institutions across the globe and uploaded to the Happywhale platform [4]. HBWs can be identi ed by their ns and special marks. The WLT dataset consists of 1 770 images collected by citizen scientists in South Africa. They were either uploaded to iSpot [7], a citizen science project which collects images or sent to the WLT project, a conservation project sta ed by volunteers. WLTs can be identi ed by their unique spots. One part of this dataset consists of 164 labelled individuals comprising 430 images and an unlabelled proportion comprising 1 340 images. The model developed in this paper consists of two main components, an object detection model and a matching classi er model. In some images, the ? Acknowledgement to AIMS South Africa for the Research Masters scholarship, CHPC for computational resources, Dr. John Measey and Mr. Alex Rebelo for collecting and providing the toad photographs.
E. Kabuga et al.
animal only occupies a small region of the image. This makes the
individualspeci c marks { spots for WLTs, and tail ns for HBWs to not be clearly visible
while they are the key feature of this study. As a result, the goal of the object
detection model is to { detect the region of the image containing the animal,
localise it using a bounding box, and extract the animal, which is then taken
to the next level of photo-matching. The object detection model is a custom
convolutional neural network (CNN) originally inspired from VGG16 [
The detection model achieved reliable results on both datasets for the task of localising the region of the image containing the animal on both datasets. The model achieved the intersection over union (IoU) of 0.90 and the coe cient of determination R2 of 0.91 for HBWs and the IoU of 0.86 and R2 of 0.85 for WLTs. The Siamese network model results are good for both HBWs and WLTs. The model correctly identi ed if a pair of images is from the same individual or not respectively in 95% of cases for HBWs and 87% of cases for WLTs. The main di erence in the performance is due to the di erent amount of data used to train the model. In this study, the semi-supervised approach on WLT unlabelled dataset has been partially successful. The model was able to identify 47 new matches from 26 individuals comprising 63 images. These identi ed matches seem to be relatively few in numbers. Without an exhaustive check of the data, it is not clear whether this is due to the failure of the semi-supervised approach, or because there are not many matches in the data. After adding the newly identi ed and labelled individuals to the WLT labelled dataset, the model slightly improved its performance and correctly identi ed 89% of WLT pairs. WildID achieved good results on WLTs compared to HBWs. It ranked the true match in the rst position in 64% of cases for WLTs and 36% of the cases for HBWs.
The Siamese network model achieved good results on the individual identication task for species dotted with individual-speci c marks. Its performance was very competitive compared with WildID.
Using neural networks to identify individual animals from photographs