Anime Characters Generation with Generative Adversarial Networks
Shengrui Ruan · 2022 IEEE International Conference on Advances in Electrical Engineering and Computer Applications (AEECA) · 2022
Automatic anime character generation has the potential to motivate professionals to create new characters while also lowering the cost of creating animation. Adversarial generative networks (GANs) have yielded impressive results in the field of image synthesis. There have been some attempts to apply the GAN model to produce anime character facial images. The increasing number of contributions in recent years requires a systematic summary and analysis of new findings to speed up future research. This paper presents a comprehensive analysis of the GAN-based generation of anime characters. This paper introduces the concept of generative adversarial networks at first, including the adversarial idea as well as its algorithms. Then, the author presents GANs’ application in anime character generation, including the ideas and new findings in this field. Finally, the author also offers a summary of the open challenges for anime character generation.