Advancements and technical challenges in generative models of artificial intelligence painting

Abstract. Artificial Intelligence Painting, as a frontier field of AI technology application, has gained wide attention and rapid development in recent years. With the advancement of generative models, especially the emergence of variational autoencoders (VAEs), generative adversarial networks (GANs), and diffusion models, the ability of AI in art creation has significantly increased. These techniques not only enable computers to generate realistic images, but also to simulate complex artistic styles, opening up new possibilities for artistic creation. Nevertheless, how to maintain a high degree of complexity and realism in the generated images, how to improve the data quality to enhance the generalization ability of the model, how to solve the pattern collapse problem to improve the stability of the generation, and how to improve the efficiency while ensuring the quality of the generation are all key issues in the current research. In addition, with the continuous evolution of AI technology, the future development direction and potential applications are also worth exploring in depth. The paper aims to systematically analyze the generative models and their applications in AI painting, to deeply discuss the technical challenges faced, and to look forward to the future development prospects.