Given two views of the same scene, could the output of something like (NotActuallyOpen)Pose theoretically be combined to create a marker-less 3D motion tracking

Given two views of the same scene, could the output of something like (NotActuallyOpen)Pose theoretically be combined to create a marker-less 3D motion tracking solution?