**The growth engine of the IT industry, ‘virtual human’… Competition for technology development**
Recently, as the metaverse and virtual reality (AR) markets expand, virtual humans, or ‘virtual humans’, are attracting attention. Currently, virtual humans are active in the financial industry, entertainment, and e-commerce fields, and are active in various fields of society.
In Korea, as the first virtual influencer in Korea, ‘Rosie’, which gained great popularity as a Shinhan Life advertisement model last year, is working as a virtual human model for ‘Sua’ of Kakao and ‘Lucy’ of Lotte Home Shopping. Recently, Naver announced that it will be active in various Naver contents in the future by unveiling ‘Isol’, a virtual human for the first time.
In accordance with this trend, it can be seen that various domestic and foreign companies are working on virtual human creation and technology development, and are using it in various marketing fields such as bankers, anchors, and online class instructors.
All currently active virtual influencers and virtual human models are collectively referred to as ‘virtual humans’, but different technologies are applied depending on the purpose of use.
1. Digital Double
It is a technology that takes a video of a real person and then puts on a virtual face through computer graphics (CG) work. Through ‘3D modeling’, Digital Double three-dimensionally analyzes the shape of a person’s face with hundreds of 3D cameras. Then, through ‘structural analysis’, facial muscle movements are analyzed so that they can move and make expressions like a real person. Unlike deepfake (fake video) technology, which stitches 2D shape images, it directly shoots an existential model and analyzes the shape, so a face shape can be realized closer to the real shape.
Digital double technology is mainly used in advertising and media fields such as Instagram and YouTube that require a lot of movement or performance. It is also a representative technique applied to Rosie and Lucy introduced earlier.
2. Text-to-Speech Synthesis (TTS)
This is a technology that creates the voice of a specific person by inputting a word or sentence. When a real human voice is recorded, the human voice, which is analog data, is converted into digital data. After that, it is taught by AI to learn habits and intonations. Because AI learns the relationship between text (words/sentences) and sound sources on its own and analyzes tone, tone, and tone of voice, it can produce a natural voice just like a real human.
Text-to-speech synthesis technology is mainly used in fields such as AI anchors, bank tellers, and online class instructors.
3. Rigging In
3D computer animation, it is a technique to create and plant a character’s skeleton and make the character appear to be moving. By subdividing elements such as joints and muscles of the 3D model, the facial muscles and body joints of a virtual human appear to move naturally.
As the rigging technology subdivides the work process, it is possible to express precise movements, and depending on how delicately it is applied, it is possible to reduce the sense of heterogeneity in the facial expressions and actions of virtual humans.
Due to the acceleration of digital transformation and technological advances, virtual humans are rapidly emerging as a recent trend. As it is used in various fields, the uses of virtual humans are limitless. Many companies are already choosing virtual humans as a new business or growth engine, and they are concentrating on advancement of virtual human development capabilities and technology to implement a highly realistic metaverse platform in the future.
https://www.codingworldnews.com/news/articleView.html?idxno=10457