A method is performed at a system that comprises one or more video cameras and a remote server system. The method includes obtaining, via a video camera of the one or more video cameras, a continuous stream of video data for a scene. The video data stream comprises color video data in accordance with a determination that the scene has illumination above an illumination threshold and comprises infrared (IR) video data in accordance with a determination that the scene does not have illumination above the illumination threshold. The method includes colorizing the IR video data based on a subset of the color video data. The method further includes presenting the colorized video data to a user in real time.