Accelerometers, Ground Truthing, and Supervised Learning
Accelerometers are sensitive to movement and the lack of it. They are not sentient and must recognise animal behaviour based on a human observer’s cognition. Therefore, remote recognition of behaviour using accelerometers requires ground truth data which is based on human observation or knowledge. The need for validated behavioural information and for automating the analysis of the vast amounts of data collected today, have resulted in many studies opting for supervised machine learning approaches.
In such approaches, the process of ground truthing involves time-synchronising acceleration signals with simultaneously recorded video, having an animal behaviour expert create an ethogram, and then annotate the video according to this ethogram. This links the recorded acceleration signal to the stream of observed animal behaviours that produced it. After this, acceleration signals are chopped up into finite sections of pre-set size (e.g. two seconds), called windows. From acceleration data within windows, quantities called ‘features’ are engineered with the aim of summarising characteristics of the acceleration signal. Typically, ~15-20 features are computed. Good features will have similar values for the same behaviour, and different values for different behaviours.
Researchers from EPFL and the University of Zurich have developed a model that uses data from sensors worn by meerkats to gain a more detailed picture of how animals behave in the wild.
Advancement in sensor technologies has meant that field biologists are now collecting a growing mass of ever more precise data on animal behaviour. Yet there is currently no standardised method for determining exactly how to interpret these signals. Take meerkats, for instance. A signal that the animal is active could mean that it is moving; alternatively, it could indicate that it is digging in search of its favourite prey, scorpions. Likewise, an immobile meerkat could be resting – or keeping watch.
In an effort to answer these questions, researchers from EPFL’s School of Engineering Laboratory of Movement Analysis and Measurement (LMAM) teamed up with colleagues from the University of Zurich’s Population Ecology Research Group to develop a behavior recognition model. The research was conducted in affiliation with the Kalahari Research Centre. Continue reading →
Animal biologging is a technique that’s quickly becoming popular in many cross-disciplinary fields. The main aim of the method is to record aspects of an animal’s behaviour and movement, alongside the bio-physical conditions they encounter, by attaching miniaturised devices to it. In marine ecosystems, the information from these devices can be used not only to learn how we can protect animals, many of whom are particularly vulnerable to disturbance (e.g. large fish, marine mammals, seabirds and turtles), but also more about the environments they inhabit.
Challenges when Tracking Marine Animals
Many marine animals have incredibly large ranges, travelling 1000s of kilometres. A huge advantage of biologging technologies is the ability to track an individual remotely throughout its range. For animals that dive, information on sub-surface behaviour can be obtained too. This information can then be retrieved when an animal returns to a set location. If this isn’t possible (e.g. individuals that make trips that are too long or die at sea), carefully constructed summaries can be relayed via satellite. This option provides information in real time, which can be very useful for researchers.
Tracks of juvenile southern elephant seals. Red tracks are individuals that returned to their natal colony. Grey are those individuals whose information would have been lost had it not been transmitted via the Argos satellite system.
Around the world there are concerns over the impacts of land use change and the developments (such as wind farms). These concerns have led to the implementation of tracking studies to better understand movement patterns of animals. Such studies have provided a wealth of high-resolution data and opportunities to explore sophisticated statistical methods for analysis of animal behaviour.
We use accelerometer data from aerial (Verreaux’s eagle in South Africa) and marine (blacktip reef shark in Hawai’i) systems to demonstrate the use of hidden Markov models (HMMs) in providing quantitative measures of behaviour. HMMs work really well for analysing animal accelerometer data because they account for serial autocorrelation in data. They allow for inferences to be made about relative activity and behaviour when animals cannot be directly observed too, which is very important.
In addition to this, HMMs provide data-driven estimates of the underlying distributions of the acceleration metrics – and the probability of switching between states – possibly as a function of covariates. The framework that we provide in ‘Analysis of animal accelerometer data using hidden Markov models‘ can be applied to a wide range of activity data. It opens up exciting opportunities for understanding drivers of individual animal behaviour.
Verreaux’s eagles lay one to two eggs but only raise one chick. These are normally hatch during mid-winter and stay on the nest for around 90 days. The parents provision the chick during this time, delivering prey to the nest. Rock hyrax is a frequent item on the menu, as seen here. Having concurred data from accelerometers and nest cameras for the same bird could allow for activity level to be linked to successful prey acquisition or prey type. (Photo taken by a nest camera installed as part of the Black Eagle Project, Megan Murgatroyd)
Fish and invertebrates predominantly or exclusively detect particle motion.
A growing number of studies on the behaviour of aquatic animals are revealing the importance of underwater sound, yet these studies typically overlook the component of sound sensed by most species: particle motion. In response, researchers from the Universities of Bristol, Exeter and Leiden and CEFAS have developed a user-friendly introduction to particle motion, explaining how and when it ought to be measured, and provide open-access analytical tools to maximise its uptake. Continue reading →