Tencent Cloud AI Digital Human
- Product Overview
- Digital Human Platform Operation Guide
- Avatar Production and Asset Management
- Digital Human Conversation Interaction Application and Management
- Introduction of Avatar
- Basic Image Library
- Guide on Avatar and Voice Clone
- Server API Integration
- Avatar Image Customization and Voice Clone API Documentation
- Avatar Customization
- Voice Clone (Basic Edition)
- Voice Clone (Ultra Edition)
- Video Generation Service API Documentation
- Interactive Digital Human Service API Documentation
- Live Session Management
- Instruction-driven
- Other APIs
- Personal Asset Management API Documentation
- Client SDK Integration
- H5 SDK Integration
- Client Rendering API Integration
Introduction to Image Categories
Last updated: 2025-03-13 17:06:20
Introduction to Image Categories
Avatar Type | Definition | Use Cases | Example |
2D Premium | By recording motion materials in a professional studio, a digital human applied to broadcasting and interactive scenarios can be generated after about two weeks of training. The boutique image can randomly insert specified motions in the text, and the motions are diverse. | Applicable to customers in finance and media who have requirements for the image and motion of digital humans. |  Loading… |
2D small sample - general lip movement | A digital human is trained through a real-person video footage. The appearance of the digital human is consistent with that of the real person, and the mouth shape will use the general lip and teeth generated by the large model. The requirements for training video footage are lower. For details, see Image Recording Guide - General Mouth Shape. | Applicable to customers who have no requirements for the mouth shape of digital humans and no good shooting conditions. |  Loading… |
2D small sample - exclusive mouth shape | A digital human is trained through a real-person video footage. The appearance of the digital human is consistent with that of the real person, and the mouth shape will use the exclusive lip and teeth of the real person. The training video footage should have no other voices or obvious environmental sounds. For details, see Image Recording Guide - Exclusive Mouth Shape. | Applicable to customers who have requirements for the image replication of digital humans and good shooting conditions. | |
2D small sample - high precision version | A digital human is trained through a 4K real-person video footage. The material collection requirements and the final lip and teeth effect are the same as those of 2D small sample (exclusive mouth shape). The resolution of the final digital human is upgraded to 4K. For details, see Image Recording Guide - High Precision Version. | Applicable to large conferences, face-to-face dialogues, product launch events, large screen scenarios. | |
2D small sample - photo avatar | A digital human can be trained through a photo. This version is designed for low-cost and quick turnaround. Generally, it is ready for use within 10 minutes after material submission. | Applicable to pan-internet and entertainment scenarios. |  Loading… |
3D Cartoon | Set the facial features, hairstyle, clothing, accessories, etc. of the digital human according to the customer's requirements to complete the original painting. After the customer reviews and finalizes the final image, proceed with model making. After stages such as bone binding, rendering, and UE tuning, a digital human that covers interactive and broadcasting scenarios can be output. | Applicable to scenarios where there is an existing 2D mascot image and it is expected to be upgraded to a 3D image to provide services for users. |  Loading… |
3D Semi-Realistic | Set the facial features, hairstyle, clothing, accessories, etc. of the digital human according to the customer's requirements to complete the original painting. After the customer reviews and finalizes the final image, proceed with model making. After stages such as bone binding, rendering, and UE tuning, a digital human that covers interactive and broadcasting scenarios can be output. | It is suitable for scenarios requiring a certain realistic sense but not high precision, such as news reading and mobile smart customer service scenarios. |  Loading… |
3D Realistic | Set the facial features, hairstyle, clothing, accessories, etc. of the digital human according to the customer's requirements to complete the original painting. After the customer reviews and finalizes the final image, proceed with model making. After stages such as bone binding, rendering, and UE tuning, a digital human that covers interactive and broadcasting scenarios can be output. | It is suitable for scenarios requiring high realistic sense and high-precision display, such as brand promotion, large-screen interaction scenarios. |  Loading… |
Image Comparison
| 2D Small Sample - General Lip Movement | 2D Small Sample - Exclusive Mouth Shape | 2D Small Sample - High Precision Version | 2D Small Sample - Photo Avatar |
Recording requirements | Record a video of at least 1 minute. There is no requirement for the sound of video shooting. | Record a video for at least 3 minutes. The recording environment needs to be quiet, and only the sound of the subject can be recorded. | The recording standard is the same as that of exclusive lip synchronization. The video resolution must be 4K. | Only one clear front photo of a person is required. |
Delivery cycle | Deliver a demo within 1 day for customer effect confirmation. Customers can use it after clicking to confirm. | Deliver a demo within 2 days for customer effect confirmation. Customers can use it after clicking to confirm. | Deliver a demo within 3 days for customer confirmation. Customers can use it after clicking to confirm. | Available for use within 10 minutes. |
Finished product effect | The general version uses lips and teeth generated by big data models. | The exclusive version records one's own lip movement, with better facial clarity. | Based on the effect of exclusive lip synchronization, output in 4K resolution for higher definition. | The photo avatar uses lips and teeth generated by big data models, and the body pose cannot sway slightly. |
General lip movement vs Exclusive lip movement |  Loading… | | | |
General lip movement vs Photo avatar |  Loading… | | | |
Exclusive lip movement vs High-precision version |  Loading… | | | |
Was this page helpful?
You can also Contact Sales or Submit a Ticket for help.
Yes
No