The track of AI-generated videos is about to change.
In February, OpenAI's video generation model Sora was born, which made many domestic video and AI manufacturers fail to have a good year. From the outside world's point of view, overseas technology is already so mature, but no one at home seems to be able to compete with it.
In fact, the accumulation of technology by Chinese manufacturers in audio and video generation has never stopped.
On April 18, the 2024 China Generative AI Conference was held in Beijing, and Wondershare Technology (300624.SZ), an A-share listed company of AIGC software, was invited to participate in the conference, and announced that its audio and video multimedia model Wondershare "Tianmu" will be officially tested on April 28, in order to further optimize the performance and experience of "Tianmu" through extensive user participation and feedback, and explore the application potential of multimedia large model technology in different fields.
At present, the "Tianmu" model of Wondershare has been filed with the Cyberspace Administration of China, and its capabilities have covered atomic capabilities such as Wensheng video, video video, Wensheng music, video soundtrack, Wensheng sound, Wensheng diagram, and Tusheng diagram, and the relevant capabilities have been commercialized on a large scale in the products of Wondershare Technology. In terms of its Chinese student video capabilities, it has achieved the coherence of different styles, rich scenes and themes, and the one-click generation time supports 60 seconds+.
The Wondershare Sky Canopy model will be in public beta on April 28
As the first audio and video multimedia model in China, Wondershare "Tianmu" focuses on digital creative vertical creation scenarios, based on 1.5 billion user behaviors and tens of billions of localized high-quality audio and video data precipitation, and based on audio and video generative AI technology, it builds an AIGC application foundation based on large model architecture, empowers global creators in the whole link, and promotes large models to enter the 2.0 era.
In the future, nearly 100 audio and video atomic capabilities of Wondershare "Tianmu" will be fully integrated into the company's matrix products. During the public beta period, the company will focus on opening application tests in multiple scenarios in the fields of video creativity, audio creativity, and image creativity, such as Wensheng video, video video, Wensheng music, Wensheng sound effects, etc. Among them, the video-to-video function supports one-click video style conversion to make the picture more brilliant, and Wensheng music and Wensheng sound further upgrade the in-depth understanding of text, as well as the multi-dimensional integration ability to generate corresponding style audio based on content understanding, and support input text to generate realistic sound and double-speed sound effects.
Screenshot of Wondershare "Tianmu" Wensheng video "Boy's Adventure".
At the conference, Zhu Wei, vice president of Wondershare Technology, and more than 50 heavyweight representatives from industry, academia, research and investment circles, including guests from Alibaba, Tencent, and NVIDIA, as well as academic guests from Peking University and Tsinghua University, discussed on the same stage, and brought a keynote speech on "Market Insight and Landing Practice of Audio and Video Multimedia Large Model".
In the past 2023, the large model has empowered thousands of industries as an infrastructure, showing a blowout growth, bringing extraordinary changes to the industry. According to incomplete statistics, in 2023, there will be 238+ large-scale model products publicly released in China, and the competition in the large-scale model industry has truly entered the stage of "100 model wars". Among them, the vast majority of products have natural language dialogue as their main function, providing chat dialogue, creative writing, code generation and other services.
Zhu Wei pointed out that at present, large models have been commercialized in the field of text and images, but there are still challenges in the application of audio and video fields, such as lack of datasets, complex video content structure and hierarchy, and high computing cost, and its mature application still needs a cycle. According to the research results of a16z, a well-known venture capital company in Silicon Valley, before 2023, there are no public video models on the market, but dozens of models will be born in 2023, with more than one million users worldwide. At present, the number of AI video models that have been put into use and made some progress in the market has reached 21.
At the same time, Zhu Wei said that the current AI video model is not yet mature, but with the development of the industry, the large model is accelerating from the 1.0 graphic era to the 2.0 era with audio and video multimedia as the carrier, and 2024 will usher in the year of AI video, and AI video has a great future. At present, Wondershare Technology has been deeply involved in creative software for more than 20 years, has rich accumulation in audio and video data, cross-modal and multimedia technology, and has a deeper understanding of global multimedia creators. Looking forward to the future, Wondershare Tianmu will adhere to the parallel of internal applications and external openness, gradually open API interfaces to external enterprises, and look forward to more industry partners joining to jointly build an open ecosystem of audio and video multimedia models, and work together to go to the AI era.
Zhu Wei, vice president of Wondershare Technology, gave a keynote speech on "Audio and Video Multimedia Large Model Market Insight and Landing Practice".
With the acceleration of digital transformation and the widespread application of AI technology, the multimedia content creation market is growing rapidly. According to public information, at present, there are 305 million video creators in the world, 4.3 billion video coverage groups, and more than 20 billion video playbacks per day, and the demand for videos is very large. According to the latest research by QYResearch, the global professional audio and video system market is expected to reach $363.27 billion in 2029, with a compound annual growth rate of 5.0% in the next few years.
Wondershare Technology, which launched Wondershare "Tianmu", is an A-share listed company with a wide product coverage, large revenue volume and high degree of globalization in the field of digital creative software in China, and is also a leading enterprise in the domestic AIGC industry, with business covering more than 200 countries and regions, with a cumulative global user base of more than 1.5 billion and a monthly active user of nearly 100 million, and is regarded as "China's version of Adobe".
Since its establishment 20 years ago, Wondershare Technology has been keeping up with the wave of technology, keeping pace with the times, gaining insight into market demand, and empowering content creation at different stages. Facing the new AIGC era, Wondershare Technology has anchored the digital creative creation scene and has launched new AIGC creative products such as Wondershare Aihua, Wondershare Broadcast, Wondershare Kwicut, Wondershare VirtuLook, etc. Ink knife and other products have been integrated with AIGC capabilities, and continue to optimize and innovate, using AI to make the world more creative.
Leifeng.com