Image Aesthetics Assessment for Virtual Cinematography of Cloud-based Performing Arts Scenes

Cloud-based performing arts is an innovative form of traditional performing arts that embraces new artistic expression. Lens language is widely used in the production of performing arts programs, which helps to enhance the artistic expression of performance content. One of the cores of its aesthetics expression is image composition. However, great majority current intelligent cinematography methods tend to overlook the learning of lens language attributes, resulting in the inability to obtain camera images that effectively reflect the composition and cinematographic intention. Therefore, in the cloud-based performing arts scenes, we propose an aesthetics assessment framework for virtual cinematography, which is designed to select the optimal camera position and angle to make its view image composition closer to the designed composition patterns until aesthetics assessment results meet expectations.