Neural Style Transfer to Design Drapes

Neural style transfer is a concept where the style of an abstract artwork is transferred to an image. Since past two years, neural style transfer has become a gravitated topic both in academic literature and industrial applications. Receiving an increasing attention from computer vision researchers, several ways are being proposed to either improve or extend the original algorithm proposed by Gatys et al. The algorithm could be applied to wide range of images and has concocted sundry applications. We devise an approach to apply our algorithm based on Gatys on designing of drapes. Importantly, this model permits a user to use his own content and style image. It provides user to enter multiple styles on a single content image. Also, a user can apply masking to provide the user with two different styles in the different regions of the content image. The output will provide a picturesque image.