Stereoscopic imaging is performed using a mobile computing device having front- and rear-facing cameras and mounted on a rotational mechanism. The cameras and rotational mechanism are controlled, e.g., by a downloaded application, to capture images of a room or other environment using both the front- and rear-facing cameras at different pan angle and tilt angle combinations. First and second images of a portion of the room or other environment, captured using the front- and rear-facing cameras, are selected for inclusion in a stereo image pair. Obtained images and corresponding metadata are transferred to a remote system. A structure from motion pipeline is used to generate a three-dimensional model of the room or other environment. Data that enables the mobile computing device to display a three-dimensional model of the room or other environment is received from the remote system and used to display the three-dimensional model.