Digital Great Wall Project

We have digitally restored the Xifengkou Section of the Great Wall at a high-precision level of
millimeters in an immersive and interactive way, being the first around inthe world to apply cloud gaming
technology to cultural relics protection.

Learn more

Digital Dunhuang & Tencent Interactive Entertainment
Cultural Heritage Digital Creative Technology Union
Laboratory

We have built a digital replica of the scripture cave and created an official digital human Jiayao for Mogao Grotto at Dunhuang

Learn more

Beijing Digital Axis – A "Digital Axis" Project for
World Heritage Application

Leveraging game technologies, we have reproduced a high-precision, interactive "Digital Axis"
to help Beijing Central Axis to apply for the World Heritage Sites

Learn more

Game Technology

Tencent Games always focuses on the breakthrough and application of
new technologies to enrich game experience.

Digital Human

xLab

Self-developed Engine

Visual Technology

Voice Technology

Cloud Gaming Technology

Game Database

In 2017, NExT Studios initiated an international joint study of high-fidelity digital humans. Gradually, it explored a self-developed path for generating digital humans. With two constant iterative technical pipelines, which are xFaceBuilder_® Digital Role Generation Pipeline and xMoCap_®3A Level Motion Capture Animation Whole Process Pipeline, NExT Studios is committed to the research into the field of digital human generation.

Siren

Siren is a high-fidelity, real-time, interactive digital human developed by NExT Studios jointly with leading companies in various fields of the industry, including Epic Games, Cubic Motion and 3Lateral. Siren's avatar is captured from a real-life woman, but Siren can be driven by anyone else. Major breakthroughs in the authenticity of details of Siren's appearance have been made as the details of her skin, the rotation of her eyes, the manner of speaking, and even the fine lines on her faces can be seen clearly. Skin texture such as pores and fine lines, as well as the slightest facial expression, is exactly the same as that of a real human.

Learn more

Matt AI

Matt AI aims to push the boundaries of using AI technology implementation to drive high-fidelity digital human. It provides a simple and effective deep learning-based approach to generate almost identical facial animations through voice and facial motion capture. Thus, the vivid and exquisite facial animation of digital humans can express a real human's subtle emotions more accurately. Matt AI debuted in SIGGRAPH Asia 2019 Real-Time Live, introducing itself to the world brilliantly.

Learn more

Steven

The high-fidelity digital double of Tencent Senior Vice President Steven Ma was developed in-house by NExT Studios. We've established the first successful film-grade production process and are constantly striving to create high-fidelity 3D content that "blurs the line between virtual and real worlds." With the help of NExT's own xFaceBuilder_® digital character production pipeline, we produced high-quality pore-level digital human skin and millimeter-level garments. We made over 700 facial expressions to make the digital human look even more realistic and used motion capture and manual editing to create the character movements. HDR restoration and shot-by-shot lighting fine-tuning enabled us to achieve a complete integration of the real scene and virtual character.

Learn more

xLab is short for Motion Capture Lab and Photogrammetry Lab under NExT Studios. The two labs respectively correspond to the self-developed xMoCap_®3A Level Motion Capture Animation Whole Process Pipeline and xFaceBuilder_® Digital Role Generation Pipeline, which cover the hardware and software simultaneously. xLab seamlessly integrates the on-site face scanning, motion capture and post-motion capture data processing, thus realizing a one-stop process from face and body motion capture to final animation output.

NExT Motion Capture Lab

With a live capture solution at a film-and-television level, NExT Motion Capture Lab provides technical supports with its self-developed xMoCap _®3A Level Motion Capture Animation Whole Process Pipeline. XMoCap _® integrates role binding, motion capture, animation tools, asset management and motion capture database, supporting large-scale teamwork, cloud-based massive asset management and high-fidelity 3D role animation production, thus meeting the development requirements of 3A games.

Learn more

NExT Photogrammetry Lab

With a scanning environment at a photo studio level, NExT Photogrammetry Lab provides technical supports with its self-developed xFaceBuilder _® Digital Role Generation Pipeline. Through extreme algorithm optimization and development of convenient tools, XFaceBuilder _® has improved the efficiency of generating high-fidelity human face and established a whole-process high-fidelity human face generation pipeline, including modeling, binding and animation, at a film-and-television level also with flexibility.

Learn more

Dawn illumination baking system

Tencent Games CROS has independently developed Dawn, a GPU distributed baking system that realizes the industry-leading global illumination algorithm. With the latest GPU hardware ray tracing ability, Dawn is able to support multi-computer and multi-card distributed baking. It provides high-fidelity global illumination effects while improving baking efficiency by dozens of times, thus significantly boosting the iterative efficiency of game scene production.
Dawn GPU distributed baking system has been implemented in dozens of projects under study in Tencent Games, offering high-quality baking services for these projects and accelerating game production at the same time.

Superman character animation pipeline

Tencent Games CROS has independently developed Superman, a one-stop pipeline tool framework for character animation that covers art tools, engines, AI technology and motion capture capability. Superman manages the data generation link from DCC characters to animation in an industrialized mode. It is suitable for the development of all kinds of high-precision character projects, and can give full play to artists' abilities on depicting and standardizing facial expressions, emotions, mouth shapes and limbs of virtual characters at multiple dimensions, thus improving the R&D efficiency of game business.
Currently, Superman has been successfully applied to many well-known games and digital human projects, such as QQ Flying Car, Tianya Mingyue Knife, Arena Breakout, and Arena of Valor.

Fully-automatic LightMapUV unfolding tool

Fully-automatic LightMapUV Unwrapping Tool is an automatic generation tool of LightMapUV (2UV) independently developed by Tencent Games CROS, which is suitable for projects that need static illumination scenes or ambient occlusion scenes. The Tool combines computational geometry with AI technology to realize a fully-automatic process from model cutting and unwrapping to UV neatness and packing, which can save more than 90% manpower on average.
Currently, Fully-automatic LightMapUV Unwrapping Tool has been widely applied to more than 20 self-developed Tencent games, which greatly improves the production efficiency of game projects.

PCGFlow & HDA library

Based on Houdini Engine, PCG process design framework and HDA library fully deconstruct art production and process design, making art more convenient and freer to use, and projects faster, safer and more efficient to iterate.
PCG has a wide range of application scenarios, basically including all data generation scenarios. In terms of typical scenarios, there are natural landforms, vegetation, rivers, road networks, buildings, cities, villages, etc. PCG is suitable for occasions with heavy workload, high repeatability, uniform quality requirements, or repeated iterations and modifications. In a key self-developed large-scale world project, PCG takes the role of the main framework of terrain iteration, and plays different roles in many other internal and external projects.

START ENGINE

Tencent START has independently developed START ENGINE, a future-oriented native cloud gaming technology solution. Currently, START ENGINE can achieve real-time and high-precision biological rendering, high-precision and high-density plant rendering, real-time tree dynamic simulation and GPU-accelerated real-time fluid simulation. In the game field, START ENGINE brings absolute improvement to game quality by introducing native cloud gaming capabilities with higher standards, so that the picture quality of real-time rendering of the game can also reach a film-and-television level; In the film-and-television field, the super cloud computing power of START ENGINE can help producers greatly improve their effectiveness and production efficiency.

Learn more

GameMatrix

Cloud Game is one of the representative applications in the 5G era. Cloud games can be regarded as a major disruptive upgrade of traditional games, in which users can play on TV, browser or mobile phone without downloading, and can realize multi-terminal archiving and data sharing. Technically, cloud games break away from the limitation of hardware terminals and run directly on the cloud server, thus realizing the features of configuration-free, download-free and cross-terminal.
Cloud Game has been launched on Tencent Pioneer Cloud Game Platform. Cloud gaming technology has been firstly applied to the Digital Great Wall Project to realize a high-precision and immersive digital restoration of the largest cultural heritage at millimeter level, which has become another symbolic example of the innovative application of cutting-edge technology and digital technology in the field of cultural protection.

Digital Great Wall Project

Digital Dunhuang & Tencent Interactive Entertainment
Cultural Heritage Digital Creative Technology Union
Laboratory

Beijing Digital Axis – A "Digital Axis" Project for
World Heritage Application