A comparative analysis of animal behavior (e.g., male vs. female groups) has been widely used to elucidate behavior specific to one group since pre-Darwinian times. However, big data generated by new sensing technologies, e.g., GPS, makes it difficult for them to contrast group differences manually. This study introduces DeepHL, a deep learning-assisted platform for the comparative analysis of animal movement data, i.e., trajectories. This software uses a deep neural network based on an attention mechanism to automatically detect segments in trajectories that are characteristic of one group. It then highlights these segments in visualized trajectories, enabling biologists to focus on these segments, and helps them reveal the underlying meaning of the highlighted segments to facilitate formulating new hypotheses. We tested the platform on a variety of trajectories of worms, insects, mice, bears, and seabirds across a scale from millimeters to hundreds of kilometers, revealing new movement features of these animals.
Tracking animal movements such as walking is an essential task for understanding how and why animals move in an environment and respond to external stimuli. Different methods that implemented image analysis and a data logger such as GPS have been used in laboratory experiments and in field studies, respectively. Recently, animal movement patterns without stimuli have attracted an increasing attention in search for common innate characteristics underlying all of their movements. However, it is difficult to track the movements in a vast and homogeneous environment without stimuli because of space constraints in laboratories or environmental heterogeneity in the field, hindering our understanding of inherent movement patterns. Here, we applied an omnidirectional treadmill mechanism, or a servosphere, as a tool for tracking two-dimensional movements of small animals that can provide both a homogenous environment and a virtual infinite space for walking. To validate the use of our tracking system for assessment of the free-walking behavior, we compared walking patterns of individual pillbugs (Armadillidium vulgare) on the servosphere with that in two types of experimental flat arenas. Our results revealed that the walking patterns on the servosphere showed similar diffusive characteristics to those observed in the large arena simulating an open space, and we demonstrated that our mechanism provides more robust measurements of diffusive properties compared to a small arena with enclosure. Moreover, we showed that anomalous diffusion properties, including Lévy walk, can be detected from the free-walking behavior on our tracking system. Thus, our novel tracking system is useful to measure inherent movement patterns, which will contribute to the studies of movement ecology, ethology, and behavioral sciences.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.