The remarkable advancements in artificial intelligence (AI)-driven image generation technologies have brought about a profound transformation across various industries, particularly in new media, video production, and gaming. AI-generated content (AIGC) has emerged as a game-changing, cost efficient solution for companies seeking high-quality visual assets while operating within constrained budgets and having limited access to traditional human resources. Through the use of sophisticated algorithms, AIGC enables the creation of stunning visuals without relying on conventional, labor-intensive workflows. Among the most prominent techniques, diffusion models have played a pivotal role in the development of AI image generation tools, giving rise to both proprietary platforms like Midjourney and open-source alternatives such as Stable Diffusion. These technologies continue to evolve, benefiting from the collaborative contributions of global programming communities. This study focuses on advancing the capabilities of Stable Diffusion, an open source AI image generation model, to address prevalent challenges in style consistency and image quality. By integrating Python and harnessing cutting-edge AI techniques, such as DreamBooth and embedding methods, the research aims to enhance the model's ability to replicate and embed distinct artistic styles. Specifically, the study targets the unique art style of the popular mobile game "Arknights" as a training objective, applying advanced techniques to refine the system's output. The proposed approach demonstrates significant improvements over the baseline model, showcasing enhanced performance in generating style consistent anime imagery. This research contributes to the evolving landscape of AI-driven art generation, offering novel insights into the application of diffusion based technologies within creative industries. By utilizing DreamBooth and embedding for style transfer and injection, the study achieves notable efficiency, drastically reducing the time required to train a new model. Ultimately, this work paves the way for more specialized and customizable AI systems in art creation, pushing the boundaries of what AI can achieve in the realm of creative expression.

Convolutional neural networks (CNNs) need to replicate feature detectors when modeling spatial information, which reduces their efficiency. The number of replicated feature detectors or labeled training data required for such methods grows exponentially with the dimensionality of the data being used. On the other hand, space-insensitive methods are difficult to encode and express effectively due to the limitation of their rich text structures. In response to the above problems, this paper proposes a capsule network (self-attention capsule network, or SA-CapsNet) with a self-attention mechanism for text classification tasks, wherein the capsule network itself, given the feature with the symmetry hint on two ends, acts as both encoder and decoder. In order to learn long-distance dependent features in sentences and encode text information more efficiently, SA-CapsNet maps the self-attention module to the feature extraction layer of the capsule network, thereby increasing its feature extraction ability and overcoming the limitations of convolutional neural networks. In addition, in this study, in order to improve the accuracy of the model, the capsule was improved by reducing its dimension and an intermediate layer was added, enabling the model to obtain more expressive instantiation features in a given sentence. Finally, experiments were carried out on three general datasets of different sizes, namely the IMDB, MPQA, and MR datasets. The accuracy of the model on these three datasets was 84.72%, 80.31%, and 75.38%, respectively. Furthermore, compared with the benchmark algorithm, the model’s performance on these datasets was promising, with an increase in accuracy of 1.08%, 0.39%, and 1.43%, respectively. This study focused on reducing the parameters of the model for various applications, such as edge and mobile applications. The experimental results show that the accuracy is still not apparently decreased by the reduced parameters. The experimental results therefore verify the effective performance of the proposed SA-CapsNet model.