Anhui Province: An Emerging Digital Human Hub
Anhui has become a prominent hub for digital human development and applications in China, with wide-ranging initiatives spanning government services, education, and industry. The provincial government and local agencies are early adopters of AI-driven virtual humans in public services. For example, the Anhui Housing Fund system launched an AI digital human named “Ma Xiaojing” as a customer service assistant to handle citizen inquiries, demonstrating how digital employees can augment public service delivery. Similar AI digital human assistants have been rolled out in social security and other departments, providing 24/7 interactive help to residents. Local museums and tourist sites are also experimenting with virtual guides. Notably, iFlytek (a leading Anhui-based AI firm) provided the technology for a virtual tour guide “Xiaoyan” at Beijing’s Old Summer Palace (Yuanmingyuan), who can converse naturally with visitors and narrate historical stories with lifelike expressions. These initiatives highlight Anhui’s public sector embrace of digital humans to enhance user experience.
On the industry front, Anhui is home to a growing cluster of companies specializing in 2D/3D digital human production for entertainment, marketing, and virtual services. Firms such as Anhui HipHop Intelligent Technology (Xihatek), Yunban Digital Technology, and Fuliangxing Digital Technology offer onestop platforms to create virtual avatars, live-streaming hosts, and AI clones of real people. For instance, Xihatek’s “AvatarGo” system can generate a digital human from a single photo and drive it to speak in sync with any input text, all in real time and even offline. This lowers the barrier for individuals and businesses to create their own virtual idols or customer service avatars. Yunban Digital Technology, founded in 2023, provides an AI SaaS platform for virtual live-streamers and “digital human cloning” – enabling entrepreneurs to deploy AI-driven virtual influencers at scale. Yunban’s rapid growth was marked by an A-round funding in January 2024 to expand its R&D, reflecting investor confidence in Anhui’s AIGC (AI-generated content) sector. Another local startup, Evolution Tech (进化论科技), debuted an embodied emotional intelligence platform in late 2025, filling a domestic gap in full-stack “emotion AI” solutions. Their platform combines a multimodal large model with specialized modules for emotion perception, understanding, and expression, allowing digital humans to recognize 87 nuanced emotional states and respond with lifelike voice, facial expressions, and gestures. The launch of this platform, along with applications like AI companions for the elderly and mental health monitors, marked a new phase where digital humans in Anhui are evolving from simply “able to answer” to “capable of empathy,” adding a human-like warmth to human-computer interaction.
Academic and educational institutions in Anhui are deeply involved in digital human research and talent development. Universities such as Anhui University, USTC (University of Science and Technology of China in Hefei), and the Anhui Fine Arts Academy have incorporated virtual human design, animation, and AI courses into their curricula. Student teams from these universities actively participate in national competitions on virtual idol creation and AI interaction design, often winning awards that contribute to Anhui’s reputation. In 2024, the Hefei Institute of Artificial Intelligence (a research institute under the national science center) collaborated with the city hotline to develop a video sign-language digital human for the hearing-impaired. This AI sign language avatar can interpret between a deaf user and a call center operator, translating sign gestures to text or speech and vice versa in real time, enabling barrier-free communication for the deaf community. Such projects benefit from strong provincial support: Anhui’s “Digital Economy Talent Cultivation Program (2024–2027)” was launched to build a skilled workforce in AI and digital content creation, ensuring a steady pipeline of experts in virtual reality and digital human technologies. The province also incentivizes innovation through patent support – on February 13, 2025, a Hefei-based institute applied for a patent on a “multimodal digital human generation method” to advance interactive virtual humans (indicating ongoing R&D in the local AI community).
From a policy perspective, Anhui has integrated digital human development into its digital economy blueprint. The province encourages 产教融合 (industry-education integration), where companies set up joint labs with universities to work on speech synthesis, computer vision, and motion capture – all core technologies for realistic digital people. There is also a focus on intellectual property: local firms have filed patents for innovations like complex virtual environment interactions and background management systems for AI-driven live streaming, providing technical underpinnings that improve the fidelity and manageability of digital humans. With the support of government initiatives and a collaborative ecosystem, Anhui’s digital humans are finding cross-industry applications – from AI news anchors on local TV, to virtual customer assistants on e-commerce platforms, and even AI legal assistants that can offer basic legal guidance through a human-like interface. This vibrant landscape, combining government backing, academic research, and entrepreneurial vigor, has positioned Anhui as an emerging powerhouse in China’s virtual human industry.
Hefei: A Major Center of AI Virtual Humans
Within Anhui, the provincial capital Hefei has become a major center of digital human innovation, often cited as one of China’s leading “virtual human” hubs. Numerous tech companies, startups, and research institutes in Hefei are driving breakthroughs in 3D modeling, multimodal avatar generation, and embodied AI. One key player is Xinming Intelligent Technology (芯明智能), which focuses on spatial computing and AI chips for 3D sensing. Xinming’s technology enables rapid 3D reconstruction of human faces and bodies from scans or images, drastically reducing the time and cost to create highfidelity virtual models. Their systems capture fine facial micro-expressions and body motions with sub-millimeter precision, allowing virtual avatars to mimic human movements and expressions in real time. This capability to quickly digitize a person’s likeness and mannerisms feeds into many Hefei projects where realism is paramount.
Hefei’s tech firms are not just building static avatars, but also imbuing them with intelligence and personality. AiDa Jiyuan (合肥艾达纪元科技有限公司), a startup founded in early 2025 under USTC’s techtransfer reforms, is a pioneer in “synthetic reality” and digital asset creation. AiDa Jiyuan’s team, led by a USTC professor, made headlines by creating a digital human of Qian Xuesen (a famous scientist) that “interviewed” a live TV host on a 2024 science program. This digital Qian Xuesen was generated with advanced face animation and voice cloning techniques, enabling a lifelike recreation of the late scientist’s appearance and demeanor. AiDa Jiyuan has since developed multiple highprofile digital IP assets, partnering with entities like the League of Legends Pro League (for virtual esports characters) and exploring use cases in new media production and live e-commerce. The company’s success is tied to Hefei’s supportive innovation ecosystem – it benefited from a provincial policy giving researchers ownership stakes in spin-off companies, and quickly secured seed funding from venture capital to commercialize its “composite reality” engine. These efforts underscore Hefei’s strategy of translating academic AI research into real-world digital human applications.
Hefei-based companies are also integrating digital humans into everyday services and culture. In May 2025, local logistics platform Luge (维天运通) unveiled 52 AI-generated avatars of real truck drivers, a first-of-its-kind initiative to document frontline workers’ stories using digital humans. During the city’s annual “5·2 Trucker Festival” live broadcast, each truck driver’s digital twin introduced themselves (with voices and personalities cloned from the real drivers) and shared experiences from the road. This project – termed the “AI Pioneer Program” for truckers – aimed to empower blue-collar workers to “speak out” through AI, turning their oral histories into engaging digital content. It also demonstrated Hefei’s strength in mass-producing personalized virtual humans: dozens of realistic avatars were generated and animated simultaneously, showcasing the scalability of local AI platforms in handling diverse faces and voices. Another startup, Twin Universe Technology, gained attention in August 2024 by using its self-developed X-Studio engine to drive virtual human presenters that mirror live human motions. In a public demo, Twin Universe’s system captured a human actor’s movements and facial expressions and replicated them in real-time on a 3D digital avatar – effectively creating a live virtual double. The demo featured avatars representing various professions (teacher, doctor, etc.), hinting at future uses where one could teleoperate a digital human in customer service or teleeducation scenarios.
Hefei’s high-tech zones have actively attracted metaverse and digital human ventures. On August 30, 2024, three promising metaverse digital human startups – Silicon-based Intelligence (矽基智能), One Box Vision (一盒视界), and Geek Ark (极客方舟) – signed agreements to establish operations in the “Xinshi Jie” Metaverse Industrial Park in Hefei’s Xinzhan High-Tech Zone. These companies bring complementary specialties: Silicon-based Intelligence leverages its proprietary “Yandi” large language model to train digital humans and can achieve 1:1 cloning of real people for use in live streaming ecommerce and even creating virtual doctors and lawyers. One Box Vision focuses on the full pipeline of digital human production – from data acquisition and voice training to AI video generation and cloud deployment – providing turnkey solutions for businesses’ digital transformation . Geek Ark develops virtual robotic avatars and offers AI digital employees and AI customer service agents for SMEs, along with tools for managing private AI chatbots and e-commerce automation. By clustering these projects, Hefei aims to build a comprehensive metaverse supply chain. The city’s New Station (Xinzhan) High-Tech Zone has even issued a 2024–2026 Metaverse Industry Action Plan, with incentives and an innovation center to incubate such companies. This has paid off with Xinzhan now hosting industry events like AR/VR expos and hackathons, solidifying Hefei’s image as a national hotspot for metaverse and digital human development.
Crucially, Hefei is incorporating digital humans into public-facing roles to improve accessibility and engagement. Aside from the earlier example of the sign-language digital human on the city hotline, there are virtual news anchors powered by AI appearing on local media. Hefei TV, for instance, introduced an AI virtual anchor “Qianqian” who can broadcast news around the clock. Backed by iFlytek’s voice synthesis and lip-sync tech, Qianqian’s on-screen persona looks and sounds remarkably human. In fact, iFlytek revealed that its “Virtual Human Interaction Platform” has amassed a network of 468 content design partners and over 700 reusable virtual human assets, serving more than 1,000 clients by late 2022. These assets range from avatar templates to animation libraries, enabling quicker deployment of virtual hosts or brand mascots. Hefei’s courts and legal bureaus have also tested AI legal digital humans that explain legal rights to citizens via an interactive screen avatar, making legal information more approachable. And in the cultural domain, the Hefei Science and Technology Museum has a virtual guide who greets visitors and answers science FAQs in an expressive, childfriendly avatar form. Such integrations illustrate how Hefei is leveraging digital humans not just as tech demos, but as practical tools across governance, commerce, and daily life.
Behind Hefei’s rise in this field is a synergy of strong research institutions and anchor enterprises. The city hosts the Hefei Comprehensive National Science Center with dedicated AI labs, as well as key state labs at USTC focusing on speech and multimodal AI. It’s no coincidence that Hefei has produced leading AI companies (like iFlytek) that continually spin off new startups and patents. The local government actively supports this innovation cycle; for example, it has a special Science & Technology Innovation fund that co-invests in early-stage AI projects and funds research in frontier domains like embodied intelligence. As a result, many cutting-edge developments – from emotional AI platforms to sign-language AIs – have been piloted in Hefei first. By integrating these virtual humans into real use cases (from hotline services to international events like the Olympics), Hefei has positioned itself as a leading city in China’s AI and digital human race.
iFlytek: Anhui’s AI Flagship and Virtual Human Pioneer
No discussion of Anhui’s digital human landscape is complete without iFLYTEK Co., Ltd., the region’s best-known AI enterprise. Founded in 1999 and headquartered in Hefei High-Tech Zone, iFlytek is a national leader in intelligent speech and language technologies. Led by its founder and chairman Liu Qingfeng, the company has grown into a publicly listed AI giant known for innovations in speech recognition, natural language processing, and computer vision. iFlytek’s expansive new AI headquarters park (“iFlytek Town”) in Hefei – opened in 2024 – exemplifies the scale of its ambitions . The campus spans 200+ acres with 400,000 m² of R&D and office space, capable of housing 15,000 personnel. Blending open public areas with cutting-edge labs, it serves as a base for iFlytek’s numerous projects and also as an AI innovation hub open to partners (hence the “town” concept, featuring park-like amenities and even exhibition zones for the public).
iFlytek’s core focus on AI R&D has directly contributed to advances in digital humans. The company’s speech synthesis and avatar animation engines power many virtual characters in China. In late 2022, iFlytek reported that its open AI platform had 513 AI capabilities and 3.7 million developers onboard, with 1.57 million AI applications built across 14 industries including digital human tech. Building on this ecosystem, iFlytek launched a Virtual Human Interaction Platform 1.0, aiming to create digital humans that “can perceive and interact” naturally. By 2022, this platform had assembled hundreds of design studios and assets as mentioned, demonstrating iFlytek’s leadership in providing the backend for virtual anchors and assistants. A concrete example came in May 2022, when iFlytek debuted an AI virtual anchor “Qian Qian” and her twin avatar “Xixi” on the Douyin (TikTok) live streaming platform. These AI anchors hosted live commerce sessions in iFlytek’s official channel, showcasing a new AIpowered live broadcast system that could automate product presentations and customer interactions. The virtual twins spoke with realistic synthesized voices and responded to viewer comments via an AI dialogue system, illustrating how iFlytek’s technology can enable round-the-clock live streams without human presenters. This experiment foreshadowed the now popular trend of virtual influencers in China’s e-commerce streams.
A milestone achievement for iFlytek was unveiled on October 24, 2024, at the 7th World Voice Expo. The company introduced its “Spark Ultra-Human Digital Human”, a next-generation virtual human powered by iFlytek’s large language model Spark (星火). This AI-driven avatar can generate highly synchronized lip movements, facial expressions, and body gestures from text input, all informed by semantic understanding of the content. For instance, if the AI is reading news about a happy event, the Ultra-Human avatar will smile and use upbeat intonation; for serious content, it will adopt a solemn expression. This is enabled by combining Spark’s natural language prowess with multimodal synthesis – a result of iFlytek’s research in cross-modal AI. The Spark LLM itself, first released in 2023 and updated continuously (it had major version upgrades through 2024), forms the “brain” of many iFlytek digital humans, allowing them not just to recite scripted text but to engage in conversation, answer questions, and even exhibit a degree of emotional intelligence. Hefei’s broader AI strategy leans heavily on such breakthroughs. In fact, iFlytek’s Spark model and its deployment in products (from AI learning tablets to smart office assistants) are a cornerstone of Hefei’s bid to be a national AI innovation highland.
Beyond software, iFlytek has also invested in hardware for digital humans and robots. It partnered with Ubie (优必选) Robotics to integrate voice AI into humanoid service robots, blending physical embodiment with voice-interactive digital personas. In line with the concept of “embodied AI”, iFlytek opened an Embodied AI Robot Data Center in Hefei in 2025, aiming to train robots that can interact in human-like ways (the project even enrolled human-shaped robots as “students” to learn movements, akin to motion capture training). All these efforts reflect how iFlytek’s presence in Hefei has a magnetic effect: it draws suppliers, startups, and research talent to the city, creating a rich ecosystem around digital humans. The company collaborates with local universities (USTC, Hefei University of Technology, etc.) through joint labs and scholarship programs, ensuring a flow of new ideas and skilled graduates. It also works closely with government agencies; for example, iFlytek is a key participant in the “Smart City” initiatives of Hefei, deploying AI virtual assistants in municipal service apps and information kiosks.
In summary, iFlytek plays a dual role – it is both a driver of technological innovation (through platforms like Spark and the virtual human engine) and an anchor for Anhui’s AI industry, around which many smaller digital human companies orbit. Its successes have reinforced Hefei’s identity as “China Speech Valley” (a nickname for the city due to its speech tech dominance) and now more broadly as a digital human innovation hub. As iFlytek continues to push the boundaries (with goals to approach AGI and to lead in Chinese AIGC fields), Anhui and Hefei stand to remain at the forefront of the digital human revolution in China.
[Jan 2026]
2025
Dec 21, 2025: Evolution Tech (进化论科技), a Hefei startup, officially launched its Embodied Emotional Intelligence Platform at the Anhui Innovation Center. The platform integrates a multimodal emotion-aware AI model and was the first full-stack emotional AI solution in the province. It can recognize 87 types of human emotions and drive digital humans or robots to respond with appropriate facial expressions, tone, and gestures. Three flagship applications were introduced: “PalPet” digital companion, an AI eldercare assistant, and a mental health monitoring agent – marking a new era of “empathetic” AI in Anhui.
Aug 2025: Hefei’s Xinzhan High-Tech Zone reported rapid progress in building a metaverse industry cluster. Since the previous year’s action plan, the zone attracted 10+ AR/VR and digital human enterprises, with a dedicated Metaverse Innovation Center and incubator spaces reaching full occupancy.
July 2025: Hefei-based startup AiDa Jiyuan, founded in Jan 2025, announced it had successfully created “Digital Qian Xuesen,” a virtual replica of the late rocket scientist. The digital Qian was featured on a CCTV science special, interacting with a live host via a screen, which demonstrated the startup’s cutting-edge face-reenactment and speech synthesis technology. This achievement stemmed from Anhui’s tech transfer policy that allowed University of Science and Technology of China (USTC) researchers, including AiDa’s founders, to commercialize their innovations.
May 2, 2025: During the 11th “5·2 Truckers’ Festival” in Hefei, logistics firm Luge (维天运通) unveiled 52 digital human avatars of real truck drivers, marking an industry-first use of AI to document frontline workers. Each trucker’s digital twin, generated via AI from their photos and voices, appeared in a live webcast to share personal stories, showcasing large-scale AI-driven content creation and giving visibility to an often-overlooked community.
Feb 13, 2025: The Hefei Artificial Intelligence & Big Data Research Institute applied for a patent titled “A Multimodal Digital Human Generation Method and System,” aiming to enhance virtual human interaction by fusing voice, facial, and gesture modalities for more natural communication.
2024
Oct 24, 2024: At the World Voice Expo in Hefei, iFlytek unveiled the “Spark Ultra-Human” digital avatar, powered by its Spark LLM. This AI avatar can produce synchronized lip movements and facial expressions in real-time according to the semantic content of speech, achieving a new level of realism in virtual news anchors and educators.
Aug 30, 2024: Three metaverse/digital human projects – Silicon-based Intelligence, One Box Vision, and Geek Ark – “group signed” to settle in Hefei Xinzhan High-Tech Zone’s Xinshi Jie Industrial Park. These projects, focusing on AI-driven virtual humans for live commerce, data-driven avatar generation, and AI digital employees, inject fresh momentum into Hefei’s metaverse ecosystem.
Aug 13, 2024: Hefei Twin Universe Technology Co., Ltd. demonstrated its X-Studio engine by showcasing virtual human presenters that mirror live human performers. The system captured a person’s motions and projected them onto a 3D digital character instantaneously, illustrating advances in real-time motion capture and avatar puppeteering.
July 29, 2024: At the Paris Olympic Games “China House” showcase, iFlytek provided AI virtual humans of Chinese athletes (e.g. swimmer Zhang Yufei and gymnast Zou Jingyuan). These lifelike avatars greeted visitors and explained Chinese sports culture in multiple languages. It highlighted how Anhui’s AI tech brought together sports and culture on an international stage.
June 19, 2024: The Hefei AI Research Institute (Anhui AI Lab) and the city’s 12345 Hotline launched a video sign-language digital human service. Deaf or hard-of-hearing citizens could video-call and communicate with a virtual sign-language interpreter avatar, which would translate sign to text for the hotline staff and voice responses back into sign language – greatly improving accessibility in public services.
May 4, 2024: Virtual Oasis (虚幻绿洲), a Hefei startup incubated by USTC, announced progress on its “emotional digital human” project. The team developed an AI virtual avatar platform with 3- hour customization (from a single photo to a rigged 3D model) and millisecond-level emotional feedback. In trials with a utility company and on Taobao live streams, their digital humans interacted naturally with users – e.g. the avatar host could detect viewer sentiments from comments and adjust its tone and facial expressions accordingly. This demonstrated a leap from purely voice-based AI to emotionally intelligent virtual agents.
May 18, 2024: Beijing’s Yuanmingyuan Park introduced “Xiaoyan,” a virtual tour guide built with iFlytek’s AI technologies. Xiaoyan, appearing as a young woman on screen, could answer tourists’ questions about the historical site with natural speech and gestures. She even leveraged a large model to enrich her explanations with anecdotes, offering an engaging guided tour experience. (This out-of-province deployment showcased Anhui’s AI prowess, as the tech was developed in Hefei.)
2022
Nov 18, 2022: At the iFlytek 1024 Developers’ Festival, iFlytek announced its AI Open Platform 2.0 had reached 513 AI capabilities, 3.7 million developers, and 1.57 million integrated applications. These capabilities – spanning speech, vision, and language – underpin many digital human solutions. iFlytek also revealed progress on its Virtual Human Interaction Platform, noting it had over 700 ready-to-use avatar assets and had served 1,000+ clients in fields like media, finance, and education. This signaled that digital human technology in Anhui was already moving from concept to widespread adoption.
May 18, 2022: iFlytek launched its AI virtual anchor “Qianqian” and her “twin sister” Xixi in a Douyin live broadcast room, debuting a new AI-powered live streaming system. The hyperrealistic virtual anchors engaged with the audience, demonstrating iFlytek’s text-to-speech, vision synthesis, and intelligent interactive capabilities in a live e-commerce scenario. This early experiment laid the groundwork for today’s virtual influencer boom and showcased Hefei’s pioneering role in AI entertainment.