AUTHOR=Sun Guanglong , Lyu Chenfei , Cai Ruolan , Yu Chencen , Sun Hao , Schriver Kenneth E. , Gao Lixia , Li Xinjian 

TITLE=DeepBhvTracking: A Novel Behavior Tracking Method for Laboratory Animals Based on Deep Learning

JOURNAL=Frontiers in Behavioral Neuroscience

VOLUME=Volume 15 - 2021

YEAR=2021

URL=https://www.frontiersin.org/journals/behavioral-neuroscience/articles/10.3389/fnbeh.2021.750894

DOI=10.3389/fnbeh.2021.750894

ISSN=1662-5153

ABSTRACT=Behavioral measurement and evaluation are broadly used to understand brain functions in neuroscience, especially for movement disorders, social deficits and mental diseases. Numerous commercial software and open-source programs have been developed for tracking the movement of laboratory animals, allowing animal behavior to be analyzed digitally. In-vivo optical imaging and electrophysiological recording in freely-behaving animals are now widely used to understand neural functions in circuits. However, it is always a challenge to accurately track an animal’s movement under some complex conditions due to uneven environment illumination, using of different animals or interference from recording devices and experimenters. To overcome these challenges, we have developed a strategy to track an animal’s movement by combining a deep learning technique, the You Only Look Once (YOLO) algorithm, with a background subtraction algorithm- DeepBhvTracking. In our method, we first train the detector using manually labeled images and a pre-trained deep-learning neural network combined with YOLO, then generate bounding boxes of the targets using the trained detector, and last track the center of the targets by calculating their centroid in the bounding box using background subtraction. Using DeepBhvTracking, movement of animals can be tracked accurately in complex environments and can be used in different behavior paradigms and for different animal models. Therefore, DeepBhvTracking can be broadly used in the studies of neuroscience, medicine and machine learning algorithms.