Artificial Intelligence 12 min read

2022 Digital Human System Basic Capability Evaluation and Observations

This report presents the background, methodology, evaluation model, results, and key observations of the 2022 digital human system basic capability assessment, highlighting technical, engineering, and security challenges, industry standards development, and future work to advance digital human technologies.

DataFunSummit
DataFunSummit
DataFunSummit
2022 Digital Human System Basic Capability Evaluation and Observations

Background : Driven by the metaverse concept, digital humans have rapidly moved from technical innovation to industrial application, with virtual anchors, customer service agents, and guides emerging. Challenges include technical gaps in 2D/3D modeling, emotional expression, intelligent interaction, and security/ethical concerns.

Standardization Efforts : The China Academy of Information and Communications Technology (CAICT) collaborated with multiple enterprises to develop international and industry standards through ITU and CCSA, releasing two standards at the 2022 Global Digital Economy Conference and drafting the "Digital Human System Basic Capability Evaluation Method" with participation from over 30 leading companies.

Evaluation Model : The assessment follows the standard and evaluates three dimensions—basic technical capability (appearance, voice, driving, interaction, content provision, persona customization), basic engineering capability (resource configuration, portability, extensibility, compatibility, timeliness, reliability), and basic security assurance capability (copyright protection, content traceability, risk control). A total of 48 test cases (27 mandatory, 21 optional) were defined.

Evaluation Process and Results : In April–May 2022, CAICT tested 10 enterprises (including Tencent, Alibaba Cloud, Huawei Cloud, Baidu, iFlytek, etc.) via on‑site visits and remote demonstrations. Nine companies passed the assessment; the overall pass rate for mandatory items was 96.61%, while optional items averaged 92.6%. Security assurance passed at a lower 81.25%.

Key Observations :

2D/3D digital human generation relies mainly on fine‑grained video recording and CG; generation from a single photo is still exploratory.

Voice interaction is universal, while visual and multimodal interaction are emerging.

Security assurance capabilities lag behind technical and engineering capabilities.

Implementation coverage varies across companies, especially in driving range, emotional expression, and digital watermark integration.

Commercial value is expanding across finance, commerce, media, telecom, entertainment, energy, etc., with diverse scenarios such as customer service, anchors, assistants, and multi‑channel deployment (Web/H5, APP, VR/AR, etc.).

Next Steps : CAICT will continue industry research, standard development, and evaluation testing, focusing on digital content generation, immersive visual standards, and the second round of capability and performance grading assessments.

Artificial IntelligenceStandardizationDigital HumanmetaverseIndustry ReportCapability Evaluation
DataFunSummit
Written by

DataFunSummit

Official account of the DataFun community, dedicated to sharing big data and AI industry summit news and speaker talks, with regular downloadable resource packs.

0 followers
Reader feedback

How this landed with the community

login Sign in to like

Rate this article

Was this worth your time?

Sign in to rate
Discussion

0 Comments

Thoughtful readers leave field notes, pushback, and hard-won operational detail here.