On March 30, Yao Tianheng, Vice President of Tencent‘s Content Platform Department, announced the launch of an intelligent creation assistant called “Zhiying”. This tool provides content creators with a range of intelligent creation tools to enhance efficiency and quality in their work.
Yao stated that content creators have always faced challenges in generating a large amount of professionally generated content (PGC) and creatively transforming user-generated content (UGC). However, the emergence of AI-generated content (AIGC) has significantly improved production efficiency and offers a viable solution to these problems. By combining AI capabilities with human creativity, AIGC enables man-machine collaboration in various aspects, such as content creation, viewpoint establishment, material generation, polishing and production.
Tencent‘s Zhiying is centered around the concept of digital humans. Users can generate a video featuring a digital human by simply inputting text or audio, which takes only a few minutes. Zhiying currently offers dozens of styles for digital humans with diverse appearances and clothing options. Creators have the freedom to choose their preferred appearance and clothing for their digital human, as well as add various actions and backgrounds to produce videos more efficiently.
Users can create their own digital avatars by uploading images and videos. These avatars have the ability to hold live-streams without any limitations of time or space. This means that users can easily perform various operations such as broadcasting, interviews, and interactions at any time without interruption.
Yao introduced three significant functions of Zhiying that revolutionized sound creation: text-to-speech, customized voice tones, and intelligent voice changing.
Zhiying’s text-to-speech function offers a wide range of voice options to cater to the dubbing requirements of diverse creators. These voices are suitable for various scenarios, including news broadcasting, short video creation, and audio novels. With Zhiying, users can complete dubbing and publishing of a 1000-word script in just two minutes. Additionally, users have the flexibility to manually adjust speech speed, homophones, and pauses as per their preference.
With Zhiying’s intelligent video technology, creators can easily edit audio and video content. The speech recognition feature automatically removes redundant words. Additionally, sound cloning and audio training technologies are available to modify specific sentences in the original video for audio rewriting purposes.
Zhiying’s voice-changing technology allows creators to transform audio into the voice of a specific person while maintaining the original rhythm. During a live demonstration, Yao showcased this feature by providing narration for Ye Wenjie, a complex and brilliant character from “The Three-Body Problem”. Zhiying had previously replicated Ye’s tone and, using its voice-changing technology, converted any recorded sound into Ye’s infectious tone.
In terms of video editing and creation, Yao pointed out that AIGC’s text-generated video technology is a revolutionary innovation that brings unprecedented opportunities and challenges to the video industry.
AIGC technology enables creators to convert their written text into video content without the hassle of collecting and processing materials. Zhiying’s segmented material presentation further streamlines the process by allowing creators to easily storyboard, add keyframes, filters, special effects, and more. This significantly reduces both the production cycle and the cost associated with video creation.