【動画】AIが作った呪術廻戦のアニメ、マジでクオリティが高すぎるｗｗｗｗｗｗｗ
An AI-generated fan animation for "Jujutsu Kaisen" is currently making massive waves online due to its unbelievably high quality.
Many viewers are astonished, mistaking it for official content, marveling at the flawless character movements, expressions, and effects.
This groundbreaking video sparks both excitement for AI's potential and crucial discussions about the future of the anime industry.
Related Keywords
AI Anime Generation
AI anime generation refers to a broad range of technologies where artificial intelligence learns from existing image and video data, text prompts, etc., to create new animations. In recent years, particularly with the development of deep learning, its quality has improved dramatically. While previously focused on generating still images, it is now becoming possible to generate smooth movements and complex effects while maintaining consistency between video frames. The evolution of this technology holds the potential for significant transformation in the anime production workflow. For example, by automating parts of the keyframe animation, in-between animation, coloring, and special effects—tasks that previously required vast amounts of time and human effort—AI could shorten production times, reduce costs, and even create an environment where creators can focus more on creative work. In the case of this "Jujutsu Kaisen" example, the AI learned the unique art style and movement patterns of the work, generating new scenes with a level of completeness comparable to, or even exceeding, that of professional animators, which is why it's praised as "too high quality." However, challenges remain, such as character continuity, complex camera work, and unintended glitches, making human correction and direction still indispensable. Nevertheless, AI is undoubtedly showing its potential to open up new possibilities for the future of animation as a creative tool.
Jujutsu Kaisen
Jujutsu Kaisen is a popular manga series by Gege Akutami, and its anime adaptation has gained immense popularity both domestically and internationally. Its appeal lies in its unique characters, meticulously crafted world, thrilling battle scenes, and stylish artwork and direction. The anime version, in particular, is highly acclaimed for MAPPA's high-quality animation and the intensity of its action scenes, captivating many fans. The choice of this work as a subject for AI anime generation holds several important implications. One is that due to its overwhelming popularity, there is a vast amount of fan art, videos, and discussions available online, providing a rich dataset for AI to learn from. AI can deeply learn the unique character designs, expressions, movement quirks, and even the effects of cursed techniques and background art styles specific to Jujutsu Kaisen from this data. Another point is that because the original manga and anime boast such high animation quality, it becomes a crucial focus to see how closely AI can approach that level. The fact that this AI-generated anime was praised as "too high quality" means that the AI has gone beyond mere imitation, capturing and reproducing the core elements of the work. This suggests that AI has evolved to a stage where it can understand specific artistic styles and aesthetics and apply them. On the other hand, AI-driven derivative works like this also raise delicate issues such as copyright and respect for creators, which will likely lead to future discussions on legal frameworks and guidelines.
Diffusion Models
Diffusion models are one of the innovative deep learning models driving the evolution of recent image and video generation AI. The basic mechanism of this model lies in the inverse idea of generating images from noise. Specifically, it first learns a "forward diffusion process" where small amounts of random noise are gradually added to clean training images until they become completely noisy. Then, it learns a "reverse diffusion process" to restore the original image by removing noise from the noisy image, much like playing the forward process in reverse. By repeatedly performing this noise removal step, it can eventually generate high-quality images according to text prompts or conditions. Prominent generative AIs like Stable Diffusion, Midjourney, and OpenAI's Sora are developed based on this Diffusion model, achieving astonishing realism and artistic quality in image and video generation. In video generation, maintaining consistent characters and backgrounds across successive frames while generating smooth movements is the biggest challenge, but Diffusion models significantly contribute to overcoming this challenge by utilizing continuity in their latent space. The AI anime for "Jujutsu Kaisen" likely also uses technology based on Diffusion models, and its ability to reproduce changes in character expressions and the dynamism of action scenes at a high level can be attributed to the model's image generation capabilities and its extension techniques for the time axis. The further development of this technology holds the potential to profoundly change the future of video content production.