Skip to main content
. 2023 May 30;17:1111908. doi: 10.3389/fnbeh.2023.1111908

Figure 4.

Figure 4

User interface for training the tracking model. The user interface for training the tracking model requires several inputs from the user. The inputs include the path to the labeled images (“ImageDir”), a JSON label file (“Label”), the number of mice in the images (“num_mouse”), the experiment name (“exp_name”), and pose pairs for defining the connections between keypoints to represent the skeleton (“pose_pairs”). Other adjustable parameters such as the learning rate (“sppe_lr,” “yolo_lr”) and batch size (“sppe_batchSize,” “yolo_batchSize”) can be modified as needed.