Virtual Digital Human refers to a virtual character with a digital appearance. It is a comprehensive product with "virtual" (existing in the non-physical world) + "digital" (created and used by computer methods such as computer graphics, graphics dyeing, motion capture , deep learning, speech synthesis and other computer methods) + "human" (having multiple human characteristics, such as appearance, human performance, interactive ability, etc.).
Virtual digital human development stage
1.Germination stage
20 In the 1980s, people began to try to introduce virtual characters into the real world, and virtual digital humans entered the budding stage. During this period, the production technology of virtual digital people was mainly hand-drawn, and the application was extremely limited.
In 1982, after the Japanese animation "Super Time and Space Fortress" was broadcast, the producer packaged the heroine Lin Mingmei as a singer who sang animated episode and produced a music album. The album successfully entered the well-known music rankings in Japan at that time. Lin Mingmei also became the world's first virtual singer.
In 1984, British man George Stone created a virtual character named Max Headroom. Max has the appearance and expressions of human beings, wearing a suit and sunglasses, and has participated in a movie and shot several advertisements. He once became a household name in the UK. Due to technical limitations, its virtual image is realized by live actors through special effects makeup and hand-painted.
2. Exploration stage
21 At the beginning of the century, traditional hand-painting was gradually replaced by technologies such as CG and motion capture, and virtual digital people entered the exploration stage.
The virtual digital people at this stage began to reach practical levels, but they are expensive and mainly appear in the film and television entertainment industry, such as digital stand-ins, virtual idols, etc.
In 2001, the character Gollum in " Lord of the Rings " was produced by CG technology and motion capture technology . These technologies were subsequently used in film productions such as " Pirates of the Caribbean " and " Rise of the Planet of the Apes ".
In 2007, Japan produced a widely recognized virtual digital person "Hatsune Miku". "Hatsune Miku" is a girl idol in the second-dimensional style. The early characters were mainly synthesized using CG technology. The characters' voices were synthesized using the VOCALOID1 series of speech of Yamaha , and the presentation was relatively rough.
3. initial stage
In 2018, the "AI Synthetic Anchor" jointly released by Xinhua News Agency and Sogou can display the image of a virtual digital person on the screen and broadcast the news after the user enters the news text, and the lip-shaped action can be synchronized in real time with the broadcast sound.
In 2019, the digital employee "Xiaopu" jointly released by Pudong Development Bank and Baidu is also a virtual digital person produced using artificial intelligence technologies such as natural language processing, voice recognition, computer vision , and can provide users with "face-to-face" banking services through mobile devices.
4. Growth stage
At present, virtual digital people are developing towards intelligence, convenience, refinement and diversification, and are entering a growth stage.
Doug Roble, head of the Digital Kingdom Software R&D Department of the American Film and Television Special Effects Company, showed his virtual digital person "DigiDoug" during his speech at TED, and he can capture and display real-time expression motions under the premise of realistic and realistic photos. STAR Labs, a subsidiary of Samsung, exhibited its virtual digital human project NEON at the CES International Consumer Electronics Exhibition. NEON is a virtual character driven by artificial intelligence . It has similar images and realistic expressions and movements, and has the ability to express emotions and communicate.
virtual digital human category
meta universe will further open up the application scenarios and imagination space of virtual digital humans: the application fields can be divided into identity (the core lies in the creation of IP and personality) and service (a virtual assistant that replaces the simple interaction of real people and a multimodal assistant that can interact deeply). It is estimated that the scale of my country's virtual digital people market will reach 70 billion by 2030, and the market size of virtual idols, virtual clones, virtual assistants, and multi-modal assistants will reach 1474/262/115/84 billion meter respectively.
virtual idol
Hatsune Miku, Luo Tianyi and other first-generation 2D virtual idols are synthesized using CG technology + Yamaha VOCALOID software to synthesize sounds, and their presentation form and commercialization scenarios continue to be enriched with technological progress;
ByteDance and Lehua Entertainment launched the virtual idol group ASOUL in 2020 and quickly became the top star;
Ranmai Technology launched the virtual hyperrealistic digital person AYAYI in 2021, and Chuangyi Technology launched the hyperrealistic virtual UP host Liu Yexi in 2021, with the main presentation forms being static pictures and videos.
Virtual clone
①Game: Game manufacturers have good virtual image technology precipitation;
②Social: Virtual social products such as SOUL and ZEPETO are very popular, creating the basic entrance and support framework of the metaverse.
③Film and TV: China's digital human special effects technology (motion capture, CG, etc.) has achieved rapid development, and it is expected that the market size of China's special effects industry is expected to exceed 10 billion.
Virtual Assistant
applies to alternative real-person services in scenarios where interaction requirements are relatively simple, such as news hosting, cultural and tourism navigation, financial customer service, education, product description, etc.; this type of product does not have high image accuracy, interaction requirements, and character creation.
Virtual Digital Human Technology Trend
Virtual Human Appearance Style Choice Two Poles: Hi-Fi + Style.
Due to the existence of the uncanny valley effect, human vision is highly sensitive to similar images, especially to the face. It is easy to fall into the "uncanny valley" in the stage of resemblance and non-image. Therefore, there are two development routes for virtual people's appearance: stylized + high-fidelity. The artistic creation space of stylized virtual characters is large, and the cartoon image makes exaggerated action expressions and will not scare the audience.
Summary
In the future, in the digital scene of the metaverse, every user needs his own virtual image, and a large number of non-user roles (NPCs) in the open world also need to achieve a thousand faces. The virtual person of Yuan Mainland digital technology "Jialan" is an evolutionary hyperrealistic digital person, as the world's first "meta-universe ecological experience officer and digital art appraiser".
Company's cooperation model for digital people includes three types. The first is the development and operation of digital people IP created by customers; the second is the hyperrealistic digital people or virtual images that the customer already has (already owns relatively complete digital assets) and handed over to Yuanshi Technology; the third is a one-stop service, where Yuandao completely customizes a digital person for customers, and is responsible for all operations and content execution.
In addition to participating in events, endorsements, and social media releases, Yuandao will also carry out specific product development for digital IP such as live streaming, cloud fashion week, brand manager, digital artwork, etc. Platform-based digital asset management and operation further realizes the value of digital assets by assetizing more digital content.
