A method for camera focus adjustment includes, at a computing system, receiving a depth image of a surrounding environment captured by a depth camera of the computing system. The depth image includes a plurality of depth pixels each encoding depth values corresponding to distances between the depth camera and objects in the surrounding environment. One or more depth pixels of the plurality of depth pixels are identified as a region of interest (ROI). Based on the depth values of the one or more depth pixels, a focus of an environmental imaging camera of the computing system is adjusted.