People's Network
People's Network >> Culture&Tourism · Sports

ICIF observation: the big model provides new opportunities for the development of cultural industry

Cai Minjie, reporter of China News Network
08:32, May 27, 2024 | Source: China News Network
Small font size

Original title: ICIF observation: big model provides new opportunities for the development of cultural industry

Open the applet, write down a few keywords, and a picture can be generated in a few seconds; Wenbo digital person with large model can be used for virtual guide and explanation, intelligent question and answer, education and science popularization, virtual image endorsement

At this ICIF, representative enterprises of large model industrial applications brought highly interactive scenes and products. Photographed by Cai Minjie

In recent years, generative artificial intelligence (AIGC) has become the focus of society and technology. The video generation technology represented by Sora expands the scope of intelligent production from text to video, which is inseparable from the support of large models. What impact will the development of the big model bring to the cultural industry?

The 20th China (Shenzhen) International Cultural Industry Expo and Trade Fair is being held here, and the representative enterprises of the application of the big model industry bring interactive scenes and products. These bring new opportunities for the development of the cultural industry.

Advertising companies use Wenshengtu technology to use image generation technology to generate materials and improve the efficiency of content production... In the exhibition area of ICIF Tencent, Tencent hybrid Wenshengtu big model attracted many visitors to interact.

"In the field of cultural creation, Tencent Hybrids is actively developing multimodal models to further strengthen the capabilities of cultural and creative graphics, cultural and creative 3D, and cultural or graphic and creative video." Liu Yuhong, vice president of Tencent Cloud and head of Tencent Hybrids big model, gave an example to say that in the face of the diverse needs of the content production industry, Tencent Hybrids is exploring digital intelligence The application of voice conversation virtual human, embodied intelligent robot and other aspects will introduce more cutting-edge technologies and applications into the field of content generation to improve production efficiency and interactive experience.

"The arrival of the multimodal big model is accelerating, and the whole content industry will be restructured, and the multimodal based applications will also show a trend of blooming." Liu Yuhong said that based on the accumulation of basic capabilities, Tencent's hybrid big model will promote the implementation of relevant applications, so that the big model can create more value. At present, more than 600 businesses and scenarios within Tencent have been accessed for testing, and the ability of Wenshengtu has been used in advertising material generation and other fields.

"Li Bai Digital Man" has learned more than one million words of Li Bai's life information based on the big model. At the same time, it has combined with the real cultural and tourism knowledge base to provide "Li Bai" with a "smart brain" and become an all-around intelligent travel guide

"Before the big model, 'digital people' had no 'smart' brain. At present, there are 'smart' digital people, and the 'brain' behind them operates in a similar way to humans. They can customize different versions according to needs, and constantly iterate products." Wan Lin, director of marketing public relations of Meta Image Technology, said that before, when there was no AI, they made large-scale 3D scenes, Manual production by the art department may take 7 to 8 working days. "After we have a new advanced AI algorithm, we can shoot videos and process them in one day to create interactive 3D scenes, which greatly improves the productivity of 3D content."

At this ICIF, the "video translation experience area" attracted attention. Photographed by Cai Minjie

In Wan Lin's opinion, at present, the overseas big models have advanced algorithm technology. China not only has fast catching up in algorithm level, but also has more rich application scenarios, which has the opportunity to achieve "curve overtaking" in application popularization.

With the development of technology such as the big model, the cultural tourism industry continues to unlock new scenes of digital cultural tourism consumption. At this ICIF, Jiadu Technology has brought a large number of solutions such as transportation model and transportation hub model. Through the docking with massive front-end business data, transportation managers and organizers can master climate change, traffic flow and other conditions. From the perspective of ordinary citizens, the big model is applied to intelligent customer service, which can provide real-time traffic consultation and question answering, and improve the travel experience of cultural tourism.

Ye Xihao, Product Director of Urban Transport Business Group of Jiadu Technology Group Co., Ltd., said that at present, the technology of large model in the mainland has been explored in many industries, and has made further progress in the application of in-depth business. "In combination with the development of culture and tourism, I think the large model can play a greater role in traffic safety and travel services in the future."

The large model can provide support for Chinese culture going to sea. In the Quwan Technology booth, the "video translation experience area" attracts attention. Its advantages are that it can achieve rapid video translation, population synchronization in the picture, etc. This technology reduces the past expensive manual translation costs and lengthy production cycle, and promotes the spread of Chinese culture in different global contexts. Deng Yin, business director of Quwan Technology, said that the company has built a full stack AI interactive technology industrial ecology, focusing on the self research and large-scale implementation of the generative 3D model, generative audio model, generative action model and dialogue model.

At present, the landing application of the big model in the cultural industry is accelerating its exploration, but it faces challenges such as the limitations of training data, and the copyright problem has not been properly solved. In this regard, some insiders suggested that a professional ethics committee should be set up to formulate ethical guidelines for content generation; Improve the technical supervision ability, and timely discover and deal with the content that infringes the copyright or does not conform to the regulations.

(Editor in charge: Yang Yubo, Li Nanhua)

Share to let more people see

Back to top