Search

Understanding the behavior of mice relies on the precise tracking of its limbs' movement. In recent years, deep learning based animal pose estimation methods can track the user-defined body points of mice from single-view camera. However, mice perform the behavior in three dimensional (3D) space, and there is still no mature device to acquire them. Hence, this paper proposed an integrated 3D behavioral tracking system of free-moving mice. This system uses four multi-view cameras to capture the behavior of mice synchronously. The automatic calibration module simplifies the most time-consuming calibration step of multi-view cameras by merging Zhang's calibration method with a movable mechanical checkboard. With the calibration parameters, the 3D frame poses of mice can be reconstructed using the deep learning pose estimation trajectories of each camera. This 3D behavioral tracking system can avoid the occlusion problem caused by single-view camera, and calibrate multi-view camera efficiently, which can provide high-precision 3D poses data of free-moving mice with less human supervision.

Understanding the connection between brain and behavior in animals requires precise monitoring of their behaviors in three-dimensional (3-D) space. However, there is no available three-dimensional behavior capture system that focuses on rodents. Here, we present MouseVenue3D, an automated and low-cost system for the efficient capture of 3-D skeleton trajectories in markerless rodents. We improved the most time-consuming step in 3-D behavior capturing by developing an automatic calibration module. Then, we validated this process in behavior recognition tasks, and showed that 3-D behavioral data achieved higher accuracy than 2-D data. Subsequently, MouseVenue3D was combined with fast high-resolution miniature two-photon microscopy for synchronous neural recording and behavioral tracking in the freely-moving mouse. Finally, we successfully decoded spontaneous neuronal activity from the 3-D behavior of mice. Our findings reveal that subtle, spontaneous behavior modules are strongly correlated with spontaneous neuronal activity patterns.

In the fields of technology such as aerospace, anomaly detection is critical to the overall system. With the large increase in data volume and dimensions, the traditional detection methods have great limitations, and thus anomaly detection algorithms based on deep learning have received widespread attention. In this paper, based on autoencoder: standard autoencoder, denoising autoencoder, and sparse autoencoder, an ensemble detection model that can extract more feature information is proposed. To make more use of these feature information, inspired by the idea of pooling layer of the CNN, two feature fusion methods are proposed. Finally, the experiment verifies that the result of this model is better than the single autoencoder model.

Animal behavior usually has a hierarchical structure and dynamics. Therefore, to understand how the neural system coordinates with behaviors, neuroscientists need a quantitative description of the hierarchical dynamics of different behaviors. However, the recent end-to-end machine-learning-based methods for behavior analysis mostly focus on recognizing behavioral identities on a static timescale or based on limited observations. These approaches usually lose rich dynamic information on cross-scale behaviors. Here, inspired by the natural structure of animal behaviors, we address this challenge by proposing a parallel and multi-layered framework to learn the hierarchical dynamics and generate an objective metric to map the behavior into the feature space. In addition, we characterize the animal 3D kinematics with our low-cost and efficient multi-view 3D animal motion-capture system. Finally, we demonstrate that this framework can monitor spontaneous behavior and automatically identify the behavioral phenotypes of the transgenic animal disease model. The extensive experiment results suggest that our framework has a wide range of applications, including animal disease model phenotyping and the relationships modeling between the neural circuits and behavior., Animal behavior usually has a hierarchical structure and dynamics. Here, the authors propose a parallel and multi-layered framework to learn the hierarchical dynamics and generate an objective metric to map the behaviour into the feature space.