Baidu AI Cloud Public Sector China Translated from Chinese

China Cultural Relics Exchange Center

Customer Service Automation

The China Cultural Relics Exchange Center and partner museums wanted a scalable way to provide narration, consultation, live streaming and cultural promotion without being limited by the time, place and staffing constraints of human guides.

2D digital-human production can be reduced to minutes, while 3D photorealistic production can be compressed to one to two weeks.
Lip-sync synthesis accuracy reached 98.5%.

Reported impact

Business outcome

The center launched a reusable virtual guide and promoter that can serve multiple museums, increase accessibility to cultural content and extend museum outreach into digital and overseas settings.

  • Customer Experience
  • Scalability
  • Speed to Market

Solution

The China Cultural Relics Exchange Center used Baidu AI Cloud's Xiling digital-human stack to launch Wenyaoyao, a virtual museum promoter designed for public explanation, live streaming and cultural outreach. The implementation is built on a platform that connects portrait driving, dialogue, voice and recommendation engines with persona management, content production and business orchestration. That allowed the team to build a culturally themed avatar that could discuss artifacts, interpret reports, guide audiences through exhibitions and serve multiple museums at the same time through digital clones. The same platform also gives operators a way to enrich the knowledge base behind the character so it can answer historical, artistic and scientific questions without traditional training cycles. Rather than treating Wenyaoyao as a standalone mascot, the project turned Baidu's digital-human platform into a reusable infrastructure layer for museum explanation, volunteer services, international promotion and future digital-IP operations.

Data flow

Museum content, artifact knowledge and event scripts are loaded into the digital-human platform, where portrait and dialogue engines render Wenyaoyao's appearance and answers and then distribute the resulting explanations and livestream sessions across museums and promotional channels.

Solution architecture

5 components · 3 layers
  1. Compute
    • Portrait Driving Engine Animates Wenyaoyao's appearance and movement for human-like presentation.
    • Intelligent Recommendation Engine Helps personalize content and service presentation within the digital-human platform.
  2. Serving
    • Digital Human Livestream Platform Lets Wenyaoyao deliver live explanations, hosting and cultural-promotion sessions.
    • Intelligent Dialogue Engine Supports question answering and conversational guidance across museum scenarios.
  3. Orchestration
    • Xiling Intelligent Digital Human Platform Provides the end-to-end production and operation platform for museum digital humans.

Architecture clues

  • Digital clones can serve dozens or hundreds of museums at the same time
  • Persona management, business orchestration and content production are on the same platform
  • The platform distinguishes between service-type and performance-type digital humans

Evidence from the source

Lip-sync synthesis accuracy has reached 98.5% (original: 数字人口型合成准确率达98.5%).
The Xiling platform connects portrait driving, dialogue, voice interaction and recommendation engines and shortens digital-human production dramatically (original: 通过平台四大引擎——包括人像驱动引擎、智能对话引擎、语音交互引擎、智能推荐引擎……原来需要一个星期做出来的2D 数字人,现在可以缩短到分钟级).
Wenyaoyao can provide explanation, guidance, consultation, hosting and livestream services to many museums at once (original: 文夭夭未来将以志愿者的身份服务于各大省市级博物馆,提供讲解、导览、咨询、主持对话、直播等服务).
Wenyaoyao was jointly created by Baidu AI Cloud, the China Cultural Relics Exchange Center and Jimu Technology as the country's first virtual museum promoter (original: 联合中国文物交流中心、极幕科技精心打造的国内首个文博虚拟宣推官“文夭夭”正式亮相).
Transparency Evidence & methodology 100/100 reference value · 89% confidence · 1 source

Review how this reference was scored and trace each normalized claim back to its documented source.

01

Scoring methodology

How scores are calculated

Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.

Customer reference value100/100
  • Problem clarity20/20
    Measures whether the customer's operational problem is specific enough to compare.
  • Implementation depth25/25
    Rewards named components, data flow, integrations, and architecture detail.
  • Outcome specificity20/20
    Measures whether the reported benefit is concrete rather than a generic claim.
  • Quantified impact25/25
    2 metrics describe measurable impact.
  • Deployment scale10/10
    Rewards evidence of a live or scaled deployment and meaningful integration scope.
Evidence confidence89%
  • Source authority20/25
    The main evidence is an official vendor-published customer story.
  • Claim traceability21/25
    10 of 12 key fields link to a named source.
  • Direct evidence20/20
    4 direct source quotes retained.
  • Date provenance10/10
    The publication date comes from the source page or publisher.
  • Source transparency8/10
    1 source reference document publisher, URL, authority, retrieval, and supported fields.
  • Implementation specificity10/10
    Rewards source-backed components, data flow, quantified impact, and a specific outcome.
02

Traceability

Sources & provenance

Field coverage shows which facts each source supports. Dates are shown only when documented by the source.

  1. Baidu AI Cloud

    China Cultural Relics Exchange Center customer story

    First-party
    https://cloud.baidu.com/customer/case/tuijianguan.html
    Source type
    Vendor-published customer story
    Published
    May 13, 2025
    Supports
    • Customer
    • Business problem
    • Solution
    • Products
    • Solution components
    • Data flow
    • Integrations
    • technical area
    • use case category
    • Business outcome
    • outcome category
    • Metrics
    • Architecture
    • evidence quotes
    • Publication date

Date evidence

  • Published on source page May 13, 2025 · Baidu AI Cloud · First-party
    Open date source
Not separately sourced Industry, Region, Company size. Treat these normalized attributes as lower-confidence until a dedicated source is added.

Related

Similar use cases

See all like this