Industry Overview - The first national standard for virtual digital humans in China, titled "Information Technology - General Technical Requirements for Customer Service Virtual Digital Humans," has been officially released, filling a gap in industry technical specifications [1] - This standard provides unified technical requirements and evaluation standards for the research, production, and application of customer service virtual digital humans, establishing a comprehensive technical specification system [1] - The standard outlines a reference framework for customer service virtual digital human systems, covering modules such as image generation, visual interaction, voice interaction, emotional interaction, image-driven operations, and maintenance [1] Market Potential - Customer service virtual digital humans are one of the most important application areas of digital human technology, currently widely used in finance, government, education, and other sectors [1] - The virtual digital human industry is transitioning from a technology exploration phase to a stage of large-scale application, with the core market size in China expected to reach 48.06 billion yuan by 2025, driving the overall industry market size to exceed 640 billion yuan [1] - With advancements in AI large model technology, digital humans are expected to evolve from "similar in appearance" to "similar in essence," with application scenarios penetrating various industries, becoming key tools for cost reduction and digital transformation [1] Company Insights - Relevant A-share concept stocks include companies such as Jiachuang Vision and Lingyun Optics [2]

Core Insights - The first national standard for virtual digital humans in China, titled "General Technical Requirements for Information Technology Customer Service Virtual Digital Humans," has been officially released, filling a gap in industry technical specifications and providing unified technical requirements and evaluation standards for the research, production, and application of customer service virtual digital humans [1][2]. Group 1: Technical Specifications - The standard establishes a comprehensive technical specification system for customer service virtual digital humans, outlining a reference framework that includes modules for image generation, visual interaction, voice interaction, emotional interaction, image driving, and operational maintenance [1][3]. - In terms of image generation, the standard mandates that 2D digital human images must ensure clear and complete facial details, while 3D hyper-realistic digital human models must have a polygon count of no less than 200,000 [1]. - The standard requires that digital humans support multimodal interactions, including voice, gestures, and body movements, and possess operational maintenance capabilities such as keyword management and material updates to ensure continuous service optimization [1]. Group 2: Performance Metrics - The national standard specifies several core performance indicators, including a lip-sync accuracy rate of no less than 90%, an average success rate of 90% for gesture interactions, and an average success rate of 90% for body movement interactions, enabling more natural body language communication [2]. - It also sets a requirement for an emotional interaction success rate of no less than 80%, allowing digital human customer service to evolve from merely responding to actually empathizing with users [2]. - To meet emotional interaction requirements, the standard mandates that digital humans must have capabilities for expression collection, posture recognition, and voice emotion analysis, enabling them to accurately assess user emotions such as joy, sadness, and anxiety [2]. Group 3: Industry Impact - The introduction of the national standard provides clear guidance for product research and optimization for domestic companies in the virtual digital human sector, injecting strong momentum into industry innovation and development [3]. - The standard is applicable for upgrading existing 2D and 3D digital human products and allows for the integration of new technologies such as artificial intelligence-generated content [3]. - A set of accompanying testing methods is being developed to provide a unified testing benchmark, helping companies quickly identify issues and optimize products [3].