Harnessing Portrait Visuals and Furniture Object for Creative Home Furnishing Design Generation

This paper presents a novel approach for generating creative home furnishing designs. The research aims to open up a new direction in the field of creative design, establishing a link between individual user attribute features, furniture object, and the automatic creative visualization. Initially, we employ portrait analysis models to extract user attributes, including age, gender, and race. By integrating data-driven methods with creative design strategies, these attributes are mapped onto crucial interior design elements, such as spatial arrangement, color palette selection, and decoration style. As a result, an initial creative interior design image is generated. Subsequently, the input furniture object is merged with the generated interior design image to produce the final creative home design image. Leveraging keyword completion technology and a text - to - image generation model, our method attains the objective of generating interior design drawings from a single user photo and a furniture object image. Our proposed method can effectively generate personalized interior design visualizations, offering significant reference for the personalization of interior design in practical applications.