A method for scene interaction includes identifying a first real scene interacting with a virtual scene and obtaining media information of the first real scene. The method also includes determining a scene feature associated with the first real scene based on a feature extraction on the media information, and mapping the scene feature associated with the first real scene to the virtual scene according to a correspondence between the virtual scene and the first real scene. Apparatus and non-transitory computer-readable storage medium counterpart embodiments are also contemplated.