Yangtze River Daily, Wuhan Client, July 16 (Reporter Li Qin, Zheng Liangzhong) "What others have not done, we have to do." From Jianghan Pass to the Yellow Crane Tower, from the Yangtze River Bridge to Tan Hualin, the light and shadow shows that swept the circle of friends made the world know Wuhan at the same time, and also made Wuhan people remember Cai Xinyuan. Now, at the age of 50, he has a new dream.
The China Optics Valley Artificial Intelligence AI Art Exhibition was held, the National University Artificial Intelligence Art Education Alliance was established, and the country's first artificial intelligence supercomputing platform for higher art education was launched...... Since September last year, Cai Xinyuan has been behind a series of AIGC (AI-Generated Content) hot events in Wuhan.
"We're moving very fast." Recently, when the reporter of the "Presence" column of the Yangtze River Daily visited, Cai Xinyuan, a professor, doctoral supervisor, director of the Art and Science Research Center of Huazhong University of Science and Technology, and the founder of Metaway Technology, said, "I have 'unrealistic ambitions' in this field and want to do something really influential." ”
Caption: A set of AI-designed costumes are unveiled, and the person wearing a black costume is Cai Xinyuan.
Sell the design made by artificial intelligence
"Breaking into the fashion circle with AI, everyone was shocked." Last month, Cai Xinyuan was invited to participate in the Greater Bay Area Forum on Artificial Intelligence and Fashion, he wrote on WeChat Moments.
Clicking on his WeChat moments, the digital light designer seems to have turned into a fashion designer, posting a large number of fashion show photos.
In September last year, the China Optics Valley Artificial Intelligence AI Art Exhibition, curated by Cai Xinyuan, was held, which was also the first large-scale artificial intelligence art exhibition in China. At the opening ceremony of the exhibition, an artificial intelligence fashion show was held at the same time, and 12 sets of fashion works created by artificial intelligence were unveiled in a centralized manner.
"With the blessing of artificial intelligence, we can quickly enter a new track." He showed the reporter of the "Presence" column the generation process of these fashion works: using hundreds of magnificent photos of Jupiter, Cai Xinyuan trained an artificial intelligence system with great creative ability for the first time in the field of cross-border fashion, and used this digital creative system to generate unique women's clothing collections again and again.
"In just one night, AI can generate 3,000 completely different designs. How to choose from 3,000 designs is the difficulty. Cai Xinyuan said that from creating something out of nothing to creating something out of nothing, we have truly entered an era of full generation, "future designers must learn to cooperate with artificial intelligence and bravely embrace this era."
More than just clothing, Cai Xinyuan is currently making breakthroughs in many fields, including interior design, architectural design, furniture design and even jewelry design. Not long ago, the relevant person in charge of a new energy vehicle brand visited, "We chatted for more than two hours, and artificial intelligence gave birth to more than 50 designs."
At present, Cai Xinyuan's team has teamed up with Aidi Group to design a set of bohemian-style clothing designed through artificial intelligence has been officially launched. "Last year we proved that AI can do design, and this year we proved that AI can sell design." In his view, the significance of artificial intelligence + art is not to show off wild skills, but to serve the landing of the industry and empower new quality productivity.
Caption: Clothing designed by artificial intelligence on the runway.
Vertical large-scale models are the absolute direction of the future
Accompanied by the crisp sound of birdsong, the ink painting scroll slowly unfolded, and a young man came into view, which is the opening scene of the first AI Chinese style comic drama "Poetry Road Life" in China. The first episode is 3 minutes, 85 shots, sound and picture are all completed by artificial intelligence, and it took 4 days.
In April this year, the upgraded version of China's first artificial intelligence art and design education supercomputing platform, ARTI Designer XL, launched by Cai Xinyuan's team, Wuhan Supercomputing Center and Wuhan Artificial Intelligence Computing Center, was launched, compared with ARTI Designer 1.0 released in December last year, the upgraded version has significantly improved image quality, image composition ability and Chinese recognition ability. "Life in Poetry" is a product created by ARTI Designer XL.
As early as 2019, Cai Xinyuan began to focus on the research and application of large models of artistic creativity, "What I did in the early stage was machine learning, and I ran in the wrong direction. With the release of ChatGPT, a large language model, generative AI has officially entered the public eye, and we have turned around in time."
According to reports, as the country's first artificial intelligence supercomputing platform for higher art education, ARTI Designer XL aims to provide students, teachers and freelance creators with more efficient and intelligent art design tools, and has covered nearly 400 universities and enterprises in China. "From the algorithm, the computing power to the calculation data are all owned." Cai Xinyuan said.
Chinoiserie is the highlight of the upgraded version. Enter "Ye Gong Hao Dragon" into the platform, and in a few seconds, an ink painting will appear on the screen, and the main characters of the picture are a teenager and a dragon in the sky. "The biggest feature of ARTI Designer XL is the full support for Chinese." Cai Xinyuan introduced that a proprietary model of Chinese aesthetics is being trained, through which Chinese culture can be better interpreted and disseminated, and Chinese culture can be promoted to the sea.
Up to now, Cai Xinyuan's team has used the platform to explore a new generation of artificial intelligence demonstration application scenarios such as AI picture books, AI exhibitions, AI metaverses, and AI light and shadow shows. Among them, the AI picture book "Great Masters · Scientist Spirit Picture Book Series" jointly created by the Yangtze River Juvenile and Children's Publishing House has been officially published, which is also the first set of artificial intelligence picture books in China.
"Models are king." Cai Xinyuan said that the "Thousand Model War" has begun, and only vertical large models are the absolute direction of the future.
Help Wuhan build an AIGC highland
In the past two hours, Cai Xinyuan mentioned the word "busy" many times, "personal time is very tight".
Following the China Optics Valley Artificial Intelligence AI Art Exhibition held in September last year, Cai Xinyuan is preparing for this year's exhibition, in addition to the art exhibition, China's first artificial intelligence art conference and the National College Student Artificial Intelligence Art Competition are advancing simultaneously, "academicians and experts will attend, the scale is very large."
Caption: Many parents take their children to visit the China Optics Valley Artificial Intelligence AI Art Exhibition.
"In the future, the integration of art and technology is a trend, which is also a development direction of contemporary art and design education." In December last year, Cai Xinyuan led the launch of the Chinese Collaborative Innovation Platform for Artificial Intelligent Art Education, and so far, more than 380 universities have participated. ”
According to reports, in addition to allowing artificial intelligence to promote the combination of production, learning and research in the art field, this innovation platform also undertakes another mission, that is, to cultivate "a new generation of artists and designers with Chinese-style modern concepts and able to control artificial intelligence technology and modern art and design methods". This year, the platform will carry out training in multiple subdivisions such as art picture books, interior design, interactive art, and animation design.
"This project not only broke the record at that time, but more importantly, it was able to look at Wuhan again with an innovative eye." As early as 2013, Cai Xinyuan led the team to complete the country's first urban light and shadow show at Jianghan Pass. For more than 10 years, he has successfully established the only national research center in China focusing on the integration of light and shadow technology, culture and technology, and at the same time, he has also reshaped the style of Wuhan with a series of brilliant light and shadow shows. Now, he wants to brand Wuhan with artificial intelligence.
"I have been exploring how to turn Wuhan into a creative hub for AIGC in the country and even the world." In his opinion, Wuhan not only has the advantage of computing resources, but also has a good research system - Wuhan Institute of Artificial Intelligence and Wuhan Institute of Artificial Intelligence of Peking University. At the same time, as a city of universities, Wuhan has a large number of students and artistic and creative practitioners, which can be enriched into the industry. In addition to talent training and industrial agglomeration, he hopes to promote more projects with exemplary and leading roles to land in Wuhan.
"We hope to help Wuhan build a highland and center of AIGC." He wrote in the WeChat circle of friends: Vows to carry out artificial intelligence art to the end.
[Editor: Xiao Pian]
[Source: Yangtze River Daily-Yangtze River Net]
Disclaimer: The copyright of this article belongs to the original author, if there is a source error or infringement of your legitimate rights and interests, you can contact us by email, we will deal with it in a timely manner. E-mail address: [email protected]