Effective graphic design is the backbone of a successful marketing campaign. It acts as a communication bridge between designers and their audience by captivating users, highlighting essential details and enhancing the visual appearance of the campaign. However, current methodologies are time-consuming and involve layer-by-layer assembly work, which requires expertise and is not easily scalable.
To address the aforementioned problem, Salesforce researchers have introduced an open source library, BannerGen, that streamlines the design process using the power of generative ai. The library consists of three parallel multimodal banner generation methods: DesignDETR, DesignInstructionsPix2Pixand Framed Stencil Recovery Adapter. Each has been trained on a large corpus of designed graph data, allowing them to speed up the design process. Additionally, all of them have been open sourced on BannerGen's GitHub repository and can be imported as Python modules, making it easy for developers to experiment with each method. BannerGen also has licensed fonts and carefully designed templates, allowing developers to create high-quality designs.
The user can upload an image from which they want to create a banner. The image is then put through a cropping process that focuses on the main elements to create multiple sub-images. Users can also specify the type of banner they want and the text they want to include. The subimages are then integrated into the selected template to create a stunning image. The final design is produced as an HTML and PNG file.
The researchers have integrated the VAEGAN framework into their approach to align the generated designs with real-world patterns. The DETR architecture has also been incorporated into BannerGen and is known as LayoutDETR. Researchers have modified the DETR decoder to handle multimodal foreground inputs. This architecture allows BannerGen to better understand background and foreground elements, resulting in better results.
BannerGen It has also incorporated InstructPix2Pix, an image-to-image editing technique powered by diffusion models. It has been adjusted to convert background images into images with superimposed text.
The third method, Frame Template RetrieveAdapter, is used to improve the diversity of generated layouts and consists of three components: the retriever, which finds the most suitable frame based on metrics; the adapter, which customizes the input images and texts to fit the frame, and the renderer which produces the layout in HTML/CSS by integrating the background layer with the user inputs.
In conclusion, BannerGen is a powerful and versatile framework that allows users to easily create custom banners by leveraging generative ai. BannerGen's architecture has been designed to learn from real designs and understand background and foreground elements. The final design is generated as an HTML and PNG file, allowing for easy manual adjustments and can be embedded into any media for immediate use. BannerGen aims to make the graphic design process less time-consuming and help users generate high-quality, professional-grade designs.
Asif Razzaq is the CEO of Marktechpost Media Inc.. As a visionary entrepreneur and engineer, Asif is committed to harnessing the potential of artificial intelligence for social good. His most recent endeavor is the launch of an ai media platform, Marktechpost, which stands out for its in-depth coverage of machine learning and deep learning news that is technically sound and easily understandable to a wide audience. The platform has more than 2 million monthly visits, which illustrates its popularity among the public.
<!– ai CONTENT END 2 –>
