Indoor Localization using Computer Vision and Visual-Inertial Odometry