/PRCWT/December 21, 2023 Shanghai - On December 8, AI start-up company Aiman Pavilion held a press conference to launch the first professional animation graphic model and AI comic tool in China to serve creators in the professional field.
This is the first AIGC large graphic model trained by massive professional animation and comic data in China, and its dataset size exceeds 5 million. The AI comic creation tool 1.0 developed based on this graphic model has a variety of practical functions, which can free comic creators from heavy work and focus more on inspiration and stories.
For the entire animation industry, this may be a watershed moment of great change. After the climb, the scenery will be completely different.
The release of Aiman Pavilion at this time coincides with the new wave.
From the concentrated explosion of big models at the beginning of the year to the emergence of practical AI tools at the end of the year, AI entrepreneurship is entering a new stage. The theme of this stage of vibrant AI products will be to liberate productivity and open up more possibilities by eliminating mechanical and repetitive labor.
At the press conference, Ai Man Pavilion CEO Yao Bizheng said that Ai Man Pavilion represents a new starting point for the application of AIGC in the comic industry, which will open the door to creativity for countless comic artists.
Small) starting point for big change
Xia Ruoshui is a cartoonist who has been practicing for more than 15 years. According to her, the creative enthusiasm of cartoonists is often exhausted during the creative process.
What many comic book readers don't realize is that the creative cycle of a comic book that takes a few seconds to read on a page lasts nearly 16 hours. This process involves five processes: storyboard, draft, line draft, color laying, and post-production. Each process is about the quality, effect, and style of a work.
And a cruel fact is that the soul part that truly carries the irreplaceability of the work lies only in the two steps of the mirror and the draft, which only account for 30% of the overall work. In other words, the inspiration and enthusiasm of the cartoonist are greatly released during one-third of the work, and in the remaining two-thirds of the time, it is gradually eliminated by the boring and repetitive physical labor.
That is, this 70% of the manual work has seriously slowed down the production cycle of comics, leaving creators, readers, and even the entire comic market firmly trapped in the productivity trap.
Like most cartoonists, Xia Ruoshui has also hired assistants to share the work, but the effect is limited. On the one hand, training assistants requires time and energy costs, and on the other hand, the work of assistants is also unstable, and the time for changing people and recruiting people will also affect the progress of the work. If you calculate a sum of money, hiring an assistant may still be a loss-making choice.
After the birth of the AI intelligent drawing generation tool Midjourney, there was a debate in the industry whether the threshold for cartoonists and designers had been removed.
After all, any creative person can enter simple text and get a stunning image in any style.
However, Mr. Xia says such generation tools are of little help to industry practitioners: "I expect the content I generate to be more professional and accurate, not more beautiful."
This is the limitation of tools such as Midjourney, and it is also the opportunity of Aiman Pavilion. Industry insiders say that to understand this "limitation" and "opportunity", one must first understand the "professionalism" required for animation creation, and the difference between Aiman Pavilion and most graphic generation models.
The first level of professionalism in the animation industry is continuity. The algorithm mechanism for generating large-scale models results in the content of its production being diffuse and random, while a comic requires the consistency of the content, style and details of the picture. This includes not only the consistency of characters and scenes, but also the overall style of the work, the style of narrative, and the processing of details, all of which require professional learning by algorithms.
In the past six months, Aiman Pavilion has carried out targeted iterative optimization to generate images in character image, scene generation, and personal style
And other aspects can be continuously unified to effectively avoid visual incongruity among readers. At present, Aiman Pavilion has been able to generate coherent and unified multi-page comics.
The second level of professionalism is refinement. Due to the limitations of algorithms, most large models of graphic generation on the market receive a long text with rich details, which will intelligently analyze the key content, and other details will drop by layer, "lazy" creation. When cartoonists design plots, no detail is useless. Strictly following the text and retaining high-precision details is to maximize the creativity and inspiration of cartoonists.
Aiman Pavilion has also made special adjustments to this. The large model will convert the generated image into text again and compare it with the original prompt word to ensure that the generated image is consistent with the original prompt word, ensure the accuracy of the details, and maximize the creativity of the cartoonist.
The third layer of professionalism is the sense of service. In the past, AI-generated pictures often had too many colors, too many details, and too rich light and shadow. When cartoonists wanted to make secondary edits, they had to pay more work. The models and tools of Aiman Pavilion pre-consider the author's secondary editing operation, and can be layered to output line drafts, color drafts, finished drafts, and so on, so that creators can add elements with more professional intentions.
It can be said that before the birth of Aiman Pavilion, it has fully considered the actual needs of creators and is committed to freeing cartoonists from heavy work, allowing them to focus more on creative works and content creation.
For the entire industry, this may be the starting point of a major change. With the help of the Aiman Pavilion graphic model and AI tools, the creative efficiency of creators is expected to increase by more than 50%.
Small) confidence and the future
Under the great change, Aiman Pavilion has its own confidence, which comes from its layout at the three levels of technology, capital and cognition.
Aiman Pavilion was established in September 2023 and is headquartered in Shanghai. Its core technology developers come from top technology manufacturers at home and abroad, and have the engineering ability to independently build large model training bases.
Among them, Wang Bin, CTO of Aiman Pavilion, has worked in Alibaba Dharma Academy, with a double doctorate from Zhejiang University and Johns Hopkins, R & D leader Chen Zonghao is the original Dharma Academy computer vision expert, and the core R & D team of Aiman Pavilion is mainly from Alibaba Dharma Academy computer vision team, more than half of which have a doctorate.
Before this start-up, since 2019, the R & D team has been deeply involved in the field of pre-training (large model), ranking at the international leading level in terms of the number of papers published in top journals and the quality and quantity of domestic and foreign patents.
The CEO of Aiman Pavilion, Yao Bizheng, has worked at Huawei and Ping An Securities, and has been deeply involved in the investment field for ten years. He has accurate industry judgment and long-term investment vision.
At present, Aiman Pavilion has received 5 million yuan of angel investment from Guomai Culture, and has reached a strategic partnership with Guomai. It can rely on Guomai's rich comic resources for model training and data set optimization. In addition, a considerable number of company team members have a background in the comic industry and have accumulated years of industry.
Through the three-level layout of technology, capital and cognition, AIGC has established a firm foothold in the field of AIGC animation creation and seized the opportunity.
The deeper the root system is, the more lush the branches and leaves will become. At present, Aiman Pavilion's professional animation graphic models and tools can be widely used in multiple links of the comic industry, penetrating the industry chain, and unlocking greater value for different users.
For professional creators, they can use Aiman Pavilion 1.0 to complete a large amount of basic work, focus more on storytelling and creativity, and consider how to enhance the content and aesthetic value of their works to gain greater creative space.
For comic fans, Aiman Pavilion 1.0, as a creative tool, can be a great inspiration for ordinary people, lowering the threshold for comic creation and serving diverse content needs.
For IP operators, Aimango 1.0 can quickly map relevant scenarios and characters, presenting a complete world view of IP. This will undoubtedly help IP operators to efficiently evaluate adaptation potential and improve IP incubation efficiency.
It is understood that Gomai has become one of the first commercial users of Aiman Pavilion, and it is expected that works created through Aiman Pavilion will be listed and released one after another within the year.
Small) starts with business, not ends with business
On january 7, 2007, in san francisco, on the big screen next to jobs, a beam of light illuminated the black apple. At this press conference, gang leader qiao released the first generation iPhone, reinventing the mobile phone. And no one knows what that beam of light means.
It wasn't until July 10, a year later, that Jobs reappeared and brought iPhone SDK to people, which is now the App Store, that the real change began to happen.
On the third day after the release, the number of applications on the iPhone SDK grew to 800, and the number of downloads exceeded 10 million times. After two months, this number became 1 billion times.
To commemorate the figure, Apple wrote a line on the official page: "Thanks a billion".
It is not the "billion" numbers that are thanked, but the developers, creators, and countless users who choose to believe in the future and are willing to touch it.
The invention of the first generation of iPhone only opened the gap in the door of the future, and the vertical application developed based on the new technology ecology truly and permanently changed the world.
If the upsurge of the big model this year has opened the iPhone moment of AI, then the vertical industry application model represented by Aiman Pavilion has begun to open the door to the future.
"Ten years ago, cartoonists were creating on paper; now, cartoonists are creating on the drawing board; in the future, I think cartoonists should create on AI tools," he said.
It can be expected that when Love Man Pavilion goes deeper into the industry, it will usher in a broader imagination.
Based on content creation, Aiman Pavilion can cooperate with more content producers and service platforms to open up the full link from content creation to productization, and apply AI technology to content fields such as comics, animation, content production, cultural film and television, thereby promoting the reform of the animation industry chain.
This kind of exploration will not only happen in the field of comic creation, but will continue to be staged in all walks of life, continuously lowering the industry threshold, and providing everyone with the ability to produce and create.
In the face of the future, humans are all children. Aiman Pavilion and more fledgling AI companies are jointly kicking off an AI revolution.
