Shulou(Shulou.com)11/24 Report--

The large AI model promotes a revolutionary change in the mode of human-computer interaction, which makes it possible for more efficient, realistic and intelligent applications of digital human beings. At the 2023 Tencent Global Digital Ecology Conference held on September 7, Tencent Cloud announced three major upgrades to its intelligent digital Homo sapiens products, relying on industry large model technology to improve the production efficiency and interactive experience of digital Homo sapiens, lower the threshold for digital human applications, and accelerate the promotion of universal benefits for AI digital people.

Among them, the newly released "universal mouth version" with a small sample of Homo sapiens, it only takes one minute for the real population to broadcast video, and the platform can use multimodal data input without training to generate a personalized mouth shape to produce a "digital Homo sapiens" similar to a real person within an hour.

Tencent Cloud Intelligent Digital Homo sapiens Creative booth

Large model technology helps Homo sapiens production to shorten the production time to 1 hour.

In April this year, Tencent Cloud Intelligent small sample Homo sapiens production platform was released to the public for the first time. The platform has the characteristics of small training samples, high production efficiency and automatic production, and can easily achieve "self-service" production of Homo sapiens. Relying on the industry model, Tencent Cloud Intelligent Digital Human Resources Factory has been upgraded to cover three major directions: higher generation efficiency, wider application scenarios, and more intelligent production and operation of 3D Digital Human beings.

In terms of generation efficiency, the newly released "universal mouth shape" version with a small number of samples of Homo sapiens, the production time and the demand for production materials are further reduced. With the blessing of AI large model technology, the general multimodal model based on self-supervision mechanism can generate personalized mouth features of each person without training by extending portrait feature extraction and mouth shape relationship binding. This promotes the reduction of the production time of 2D small sample digital people from 24 hours to 1 hour, and the production material to only 1 minute of real people broadcasting video.

In the application scene, the newly upgraded "interactive" small sample Homo sapiens supports the recognition of "listening waiting state" and can pause and change gestures according to the real conversation scene. This will make "interactive" small sample Homo sapiens is no longer limited to one-way broadcast scenarios such as short video production, but can also be widely used in interactive scenarios such as service consulting.

In production and operation, relying on the AI modeling binding algorithm, Tencent Cloud intelligently generates 3D images in the form of photos, changes the form of manual work, greatly improves the speed of 3D image generation, and supports the mass creation of 3D digital Homo sapiens; at the same time, a comprehensive upgrade of "intelligent drive" 3D digital Homo sapiens can automatically match actions for 3D digital Homo sapiens based on semantics, making the service experience of 3D "digital intelligence employees" close to that of real people. For example, around the large model customer service, digital Homo sapiens can match the corresponding expressions and actions according to the answers generated by the large model in real time, and achieve more natural interaction.

At present, Tencent Cloud Intelligent Digital Homo sapiens is widely used in oral video production, live streaming, delivery and other scenarios. In the media field, for example, technologies such as small sample Homo sapiens are injecting more AI power into the digital transformation and upgrading of media, including Southern Daily, Southern Metropolis Daily and other institutions using Digital Homo sapiens to assist the gathering and editing team to generate oral videos to further improve content production efficiency.

Launch the Digital Homo sapiens partner support Program, covering live broadcast operations, marketing agents, MCN and other areas

Previously, Tencent Cloud Intelligent put forward the direction of automating "AI + Digital Homo sapiens Factory". By building a thick aPaaS platform, Tencent Cloud aims to enable partners to apply out-of-the-box production services of Digital Homo sapiens, quickly complete a large number of customization of digital Homo sapiens image and timbre, and develop more SaaS services suitable for vertical industries and vertical scenes. At present, dozens of partners have relied on the platform to provide digital Homo sapiens live SaaS and knowledge oral SaaS applications to the industry, covering, media and financial industries.

In order to promote the wider landing of digital Homo sapiens services and enter thousands of industries with "inclusive" services, Tencent issued the "Digital Homo sapiens partner support Program" to provide digital human technical support to partners in live broadcast operations, marketing agents, MCN and other fields.

The first batch of authorized partners of Tencent Cloud Intelligent Digital Homo sapiens include Beijing Fengping Intelligent Technology Co., Ltd., Zhongshu Liancheng (Guangdong) Culture Co., Ltd., Shenzhen Shanren Niu Shang Technology Co., Ltd., and Shaanxi Yizhenuo Information Technology Co., Ltd.

The first batch of authorized partners of Tencent Cloud Intelligent Digital Scholars

Wang Qi, vice president of Tencent Cloud and head of Tencent Cloud Intelligent Solutions, said that enterprises participating in the "Digital Intelligence support Program" can not only obtain Tencent Cloud's official authorization certificate and official website display. It will also get a series of support, such as special funds for market development, joint voice of benchmarking cases, and so on.

Currently, Tencent Cloud Intelligent Digital Homo sapiens covers five image styles: 3D realism, 3D semi-realism, 3D cartoon, 2D real person and 2D cartoon, which can achieve ultra-subtle facial expressions and hundreds of body movements. support image asset management, business service configuration and content production-related services to meet the service needs in different scenarios.

In addition to the number of Homo sapiens, the application of the AI model is gradually infiltrating into all aspects of the industry, including design, research and development, production, marketing and service, and a "efficiency revolution" driven by the "big model" is quietly taking place. Based on this, Tencent Cloud starts from the actual needs of industrial customers, based on the self-developed Tencent mixed meta model, and supports the mainstream open source model applications in the industry, constantly upgrading MaaS capabilities from the three dimensions of technology base, platform capability and intelligent application, helping customers to build exclusive large models and intelligent applications, and promoting the landing application of large models in the industry. In the future, Tencent Cloud will launch more out-of-the-box large model products and services to promote the effectiveness of AI technologies such as large models in actual production and operation scenarios.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.