Abstract
Neuroscience has traditionally relied on manually observing laboratory animals in controlled environments. Researchers usually record animals behaving freely or in a restrained manner and then annotate the data manually. The manual annotation is not desirable for three reasons; (i) it is time-consuming, (ii) it is prone to human errors, and (iii) no two human annotators will 100% agree on annotation, therefore, it is not reproducible. Consequently, automated annotation for such data has gained traction because it is efficient and replicable. Usually, the automatic annotation of neuroscience data relies on computer vision and machine learning techniques. In this article, we have covered most of the approaches taken by researchers for locomotion and gesture tracking of specific laboratory animals, i.e. rodents. We have divided these papers into categories based upon the hardware they use and the software approach they take. We have also summarized their strengths and weaknesses.
Highlights
Neuroscience has found an unusual ally in the form of computer science which has strengthened and widened its scope
Once the curved whiskers are represented by straight lines, the Hough transform is used to locate them. This approach can be used a starting point by a researcher who wants to experiment with different automated background subtraction methods for gesture tracking/pose estimation but the approach is too weak itself and not robust enough to be considered for any future improvements
Another field which can benefit from automated gesture analysis is the research on understanding how neural activity controls physical activity or how the brain responds to external stimuli [143,144,145,146,147]
Summary
Neuroscience has found an unusual ally in the form of computer science which has strengthened and widened its scope. The wide availability and easy-to-use nature of video equipment have enabled neuroscientists to record large volumes of behavioral data of animals and analyze them from the neuroscience perspective. Neuroscientists would record videos of animals they wanted to study and annotate the video data manually. Even the annotation done for the same sample at different times by the same person might not be the same All of these factors have contributed to the demand for a general-purpose automated annotation approach for video data. Computer science has an answer to this problem in the form of machine learning and computer vision-based tracking methods. The research in this area is still not mature, but it is receiving a lot of attention lately. We do not restrict the review only to previous works which focus only on rodents, but we include similar approaches that could be ported to this particular case (typically other small mammals and insect monitoring applications)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
