Shiyan Li, head of the digital human and robotics business at Baidu, which created the digital model and actor, said: “Increasing demand is driving the baby boom. digital people. “In China alone, there are more than 400 million ACGN fans (animations, comics, games and novels) and a hundred billion dollar enterprise market centered on digital people. .” And according to a business registry tracking company, Qichacha, China now has more than 280,000 businesses engaged in digital people-related activities.
Another kind of digital
The launch of a digital celebrity on Baidu may not seem like much at first, as the concept of a “virtual idol” has been around for many years. For example, US virtual influencer Lil Miquela has appeared alongside real-life celebrities in online and TV commercials since 2016, attracting more than three million followers on Instagram. However, there is something different about the Chinese virtual star: a digital person with the ability to hear, speak, and interact with real people on a level never before seen. And Gong’s digital mission isn’t just limited to singing. On the latest update of Baidu App, China’s leading search plus application, Gong appears on users’ phones, helping to search and query with the real voice of the model and actress. Since this interactive search experience launched in 2021, it has seen an 18.2% increase in voice search queries on the Baidu App.
Baidu AI Cloud first started developing digital employees in 2019 in partnership with Shanghai Pudong Development Bank (SPD). They then focused their efforts on building a digital financial advisor that would provide a service equivalent to that of a human bank representative without real-life staff. Today, SPD Bank says more than 460,000 customers rely on digital people for banking and portfolio management services each month. “Access to digital people outside of regular business hours allows SPD Bank to provide 24/7 customer service with low cost and high efficiency,” the bank representative said.
More recently, a virtual anchor created by Baidu provided live commentary in sign language at the 2022 Beijing Winter Olympics to deaf viewers. In addition to looking like a real person, avatars are enhanced with speech recognition and sign language interpretation to ensure fast and highly accurate input and output. With about 430 million people around the world suffering from “disabling” hearing loss, according to World Health OrganizationThere is strong potential for this technology to be used to increase their accessibility to a wide variety of content.
XiLing: New generation on AI platform
From entertainment to public services, digital humans are set to play a larger role in our daily lives. But behind their easy and natural appearance lies a complex web of new and emerging technologies that are pushing the boundaries of AI innovation.
The Baidu AI Cloud digital celebrity and virtual sign language anchor are created through XiLing, a new digital platform launching in 2021. At the Baidu World 2022 event to be held on May 21. 6, the company announced a new capability on XiLing that supports the digital creation of live stream hosts who can sing, dance, and respond to comments in real time — without no need for a break. XiLing is unique in its ability to support the entire process of creating a digital person from creating an actual character to creating it with chat and content creation skills. One of its most prominent attributes is speed. The platform can create 3D avatars based on real people in one to two weeks, while 2D avatars can be created in just minutes.
Additionally, using XiLing’s intelligent dialogue tools, creators can quickly customize the digital human’s ability to converse, so that it adapts and learns over time. This capability is provided by Baidu’s PLATO, a dialogue model with hundreds of billions of parameters that allows digital humans to engage in open domain conversations — that is, to understand any topic and provide relevant answers. Highly accurate speech recognition and lip sync with over 98.5% accuracy enable digital humans to have smoother, more human-like interactions. “The use of advanced AI technologies will further reduce the cost of building digital people and greatly improve their interactions with real people,” Li said.
Just as every real person has their own unique skills and talents, so does the new generation of digital humans. This could even include giving digital humans the ability to be creative on their own, thanks to recent advances made by big AI models like Baidu’s. ERNIE, can generate text and create actual images when prompted. For example, digital humans are designed to act as brand spokespersons, able to independently create and post on social media, design posters, and demonstrate in videos.
