A Novel Art Intelligence System Based on Phantom Image Generation Algorithm

Nowadays, there are state-of-art representative algorithms in the research field of phantom image generation based on deep learning, but these algorithms often still have some problems, such as large time consumption and general quality, etc. In this paper, the novel art intelligence system based on the phantom image generation algorithm is studied. The innovation of the proposed model contains 2 aspects: (1) The novel style transfer algorithm is proposed with the combination of the neural network to serve as the basic step of the phantom image generation. (2) The efficient phantom image generation algorithm is designed with the special focuses on the image background blur. Then, the 2.efficient phantom image generation algorithm is designed with the special focuses on the image background blur. he experimental results show that the proposed phantom image generation produces images of the target domain with a higher similarity at the pixel level and the transfer performance is well in the visual level.