Abstract
Recent years have witnessed the substantial progress of large-scale modelsacross various domains, such as natural language processing and computervision, facilitating the expression of concrete concepts. Unlike concreteconcepts that are usually directly associated with physical objects, expressingabstract concepts through natural language requires considerable effort, whichresults from their intricate semantics and connotations. An alternativeapproach is to leverage images to convey rich visual information as asupplement. Nevertheless, existing Text-to-Image (T2I) models are primarilytrained on concrete physical objects and tend to fail to visualize abstractconcepts. Inspired by the three-layer artwork theory that identifies criticalfactors, intent, object and form during artistic creation, we propose aframework of Text-to-Image generation for Abstract Concepts (TIAC). Theabstract concept is clarified into a clear intent with a detailed definition toavoid ambiguity. LLMs then transform it into semantic-related physical objects,and the concept-dependent form is retrieved from an LLM-extracted form patternset. Information from these three aspects will be integrated to generateprompts for T2I models via LLM. Evaluation results from human assessments andour newly designed metric concept score demonstrate the effectiveness of ourframework in creating images that can sufficiently express abstract concepts.