Big Whale AI Summit · The Retail Session will be held in Beijing on April 18. 30+ CEOs/CIOs/top experts from the field of AI and retail consumption will have high-end roundtable dialogues behind closed doors throughout the day to interpret the evolution trend and landing cases of AI in the field of retail consumption.
Is a digital human a vase or not?
Big Whale AI Summit · Retail Consumption Session
AI Summit Around Big Whale · In the retail consumption session, Tiger Sniff Think Tank will launch a series of selected content in the field of retail consumption, and this is the third article. This conference covers typical cases of AI and e-commerce, AI and supply chain and other scenarios, welcome to scan the QR code to register.
The following core points are from the "AI Digital Human Landing Insight Report" written and released by Tiger Sniff Think Tank, click on the report name to get the content.
Key takeaways:
● Under the classification dimension of business and function, AI digital humans can be divided into three types: content/IP type, functional service type and virtual avatar, and the first two are applied in the field of retail consumption.
● Content/IP-based digital human is the embodiment of virtual IP/idol identity, and the landing form is presented in the form of virtual idols, digital astronauts, brand spokespersons and other products.
● Functional service-oriented products are designed to enhance and assist human work through a specific function of AI digital humans, reduce enterprise costs, and be applied in the form of virtual anchors, digital customer service, digital experts and other products, which can be uniformly regarded as digital employees.
● Retail consumer enterprises are paying more and more attention to the profitability of digital humans and the substantive effect of improving quality and efficiency, in addition, the expressiveness, rework rate, interactivity, stability and subsequent technology upgrade and iteration capabilities of digital humans are also the capability dimensions they focus on.
The needs and solutions of the first party's enterprise
A digital human is a virtual image that exists in the digital space, possessing the appearance, behavior, and characteristics of a person, and is realized by technologies such as computer graphics, motion capture, image cleaning, and AI.
From different dimensions, it can be divided into various types, from the macro level, it is divided into concrete digital human and soul digital human, which are embodied in digital content and form based on human thinking, from the dimension of driving mode, it is divided into human driven and AI-driven, which are driven by user and AIGC input and AI independent decision-making, from the rendering path, there are 2D digital humans and 3D digital humans, and according to function and commerciality, they are divided into content/IP type, functional service type and virtual avatar and other categories. This report is mainly written according to the classification of digital humans from the business and functional dimensions.
Is a digital human a vase or not?
Party A's demand for content/IP-based digital human services is mainly concentrated in three aspects: improving the effect, improving the quality and efficiency of content output, and increasing revenue.
The main logic behind it is that through high-value, surreal digital humans, it is possible to reduce the dependence on real or celebrity IPs and optimize content output. At the same time, the application of digital humans can reduce learning costs and improve output efficiency. Eventually, through advanced digital human images and background effects, more exposure can be obtained, tonality can be improved, and sales can be increased and costs can be saved.
Solutions include digital human image customization, short video generation platforms, and digital human live streaming and operation solutions. For example, the digital human customization service provided by Wondershare Technology and Zhijing Yunchuang, the short video platform of Silicon-based Intelligence, and the live broadcast solutions of Wondershare Technology and Lingxi Shenzhi are all powerful tools to promote the development of this field.
Party A's demand for functional service-oriented digital human services mainly starts from four aspects: overcoming the scarcity of anchors, reducing the labor cost of anchors, breaking the limitations of real anchors, and replacing low-end customer service repetitive work.
The core reason is that through the use of digital human services, Party A can effectively solve the problems of manpower scarcity, high cost, language and time zone limitations encountered in the field of hosting live broadcasts or providing customer service.
The solutions include digital employee image products, digital employee generation platform, digital human video creation platform, digital human live broadcast generation platform, and XR production and broadcasting solutions. For example, a series of services provided by companies such as Silicon-based Intelligence, Yuanke Vision, Jumping Intelligence, and Phase Core Technology are designed to meet these needs, so that the anchor and customer service processes in the retail consumption field can be optimized, service efficiency can be improved, costs can be reduced, and the limitations of traditional services can be broken.
Business model analysis
The cooperation model of Party A and Party B's enterprises
The main cooperation methods between technology suppliers and Party A are divided into three models: product and supporting system sales, digital human operation services, and SaaS digital human solutions and consumer software sales.
The first model focuses on creating a customized digital human avatar and a complete system for the customer, with the solution priced according to specific elements. The second model provides one-stop technical and operational services and receives a share of revenue. The last model adopts the provision of general SaaS digital human solutions, or the sale of SaaS consumption software, charging membership service fees and value-added service fees.
These three models have their own characteristics, which can meet the needs of different types of Party A and provide convenient, efficient and flexible cooperation methods.
The AI digital human capability that Party A is concerned about
When selecting content/IP-based digital human technology providers, Party A mainly focuses on four dimensions of technical capabilities: comprehensive technical strength, including rendering, interaction and technical iteration efficiency, cost reduction capabilities, covering saving labor costs, reducing computing power consumption and controlling operating costs, vertical industry experience and solution maturity, that is, the degree of product and business matching and product friendliness to application departments, and high-quality service capabilities, such as demand response speed, new demand delivery capabilities and service attitude. Through the evaluation of these dimensions, Party A can select a technical vendor with both professional and practical experience in order to effectively achieve business goals.
When Party A selects a functional service-oriented digital human technology provider, Party A also pays attention to four key dimensions: the stability and durability of AI digital human products, that is, whether the product can provide continuous, High-quality live broadcast services, the naturalness and expressiveness of AI digital humans, including sound quality, micro-expressions and true expression of motion details, the rework rate of AI digital humans after reproduction, which plays an important role in evaluating the product quality and professionalism of technology vendors, and the update iteration and follow-up operation capabilities of digital human technology determine whether technology vendors can provide Party A with continuous and cutting-edge services and solutions.
Landing challenges and suggestions
In the process of implementing content/IP-based digital human technology, it faces problems such as the performance and integration of AI digital humans, computing power requirements, high costs, and industry education.
The solution involves the collection of high-quality 3D data and large model training to improve performance, the localized deployment of models or industry cooperation to meet the demand for computing power, the scale effect of expanding application scenarios and standard products to reduce costs, and the personalized training process of target customers to reduce education costs. The application and development of digital characters is a complex system project, which requires the organic integration of technology, art and education, and continuous iteration and optimization.
The challenges of large-scale application of digital humans also encounter four aspects: technical bottlenecks, platform rules, high costs, and incomplete regulations.
The solution can start from the technical aspects of improving natural language processing technology, interactive research and development, optimizing micro-expression driver and voice rendering, and strengthening the collaboration and docking with live broadcast platforms. For the lack of a regulatory environment, the establishment of a comprehensive regulatory and real-name authentication mechanism to prevent potential risks is also an important step to ensure the large-scale promotion and application of digital humans.
AI digital human manufacturers should control customer expectations, provide differentiated services, and correctly select landing scenarios when serving enterprises. When using AI digital humans, Party A should have realistic expectations, flexible multi-platform applications, clear requirements, and maintain attention to AI technology innovation. In this way, it can not only give full play to the auxiliary role of digital humans as a tool, but also keep abreast of the latest market trends and technological trends, and enhance its market competitiveness. The key is to find efficient solutions and suitable implementation scenarios to realize the organic combination of digital human technology and business development.
epilogue
The rise of AI digital human technology has brought new possibilities to the retail consumption field. Enterprises need to understand and apply this technology to explore its value in terms of improving effectiveness, optimizing content output, and increasing revenue. At the same time, technology providers also need to provide targeted solutions to meet the needs of different Party A. However, as a new technology, it still faces various challenges in the process of landing and application, and it is necessary to solve the problems of technology, price and regulations side by side in order to truly promote the large-scale application of digital human technology.
If you want to know more about the application cases of AI digital humans in the field of retail consumption, you can sign up for the Big Whale AI Summit hosted by Tiger Sniff Think Tank on April 18 · In the retail consumption session, there will be case explanations of the application of AI digital humans in the field of retail consumption.
About Big Whale AI Summit · In the retail consumption session, we have invited representatives of a number of leading enterprises in the industry, including Yili Group, Haidilao, Bain & Company, Nature Hall Group, Baiguoyuan, Shijiufang, Shanghai Auntie, Zhuyeqing, Laobaimin Pharmacy, TATA Wooden Door, Youngor, China Chain Store & Franchise Association and other enterprises and institutions.
Is a digital human a vase or not?
We sincerely invite you to participate in the conference to discuss the latest technical practices and challenges of AI and retail consumption!
This content is the author's independent view and does not represent the position of Tiger Sniff. May not be reproduced without permission, please contact [email protected] for authorization
People who are changing and want to change the world are all in the Tiger Sniff APP
