After "AI Stefanie Sun", black technology has made a comeback.
Recently, top stream singer Taylor Swift (referred to as "unlucky" by Chinese fans) has widely circulated videos of her fluent Mandarin on various social media platforms. In the video, when Moumou speaks Chinese, he is precise and has a tone similar to that of his native speaker, even able to match his mouth shape.
Not only that, popular online videos also feature Chinese speaking Bean, Emma Watson, and English speaking Cai Ming. In addition to the exchange between Chinese and English, there are also people abroad who have tried to translate English into six other languages such as Japanese, and the results are also very good.
How did it happen?
It is understood that these videos all use the AI video tool HeyGen. The creator only needs to upload the video and select the language to be translated, and this tool can achieve automatic translation, adjust voice color, and match mouth shape.
However, HeyGen is not perfect, and the video subtitles need to be added by the creator. In addition, some creators have stated that if the video characters are deliberately compared, there is still a lack of emotional restoration when speaking another language.
This cannot stop the popularity of the tool. HeyGen has currently launched a paid version, but it can be tried out for free and generate 1 minute long videos. Some creators claim to have queued up 7000 videos to complete the translation.
Behind the scenes, players from China have achieved profitability after 7 months of product launch
HeyGen is a tool designed for video creators, and translation functionality is not all of it. At a rough glance, HeyGen has very powerful features around video creation, including host face-changing, text relay broadcasting of videos, creating virtual hosts, and voice cloning.
The creation of digital people is its main function. According to HeyGen, even without cameras, studios, and actors, ordinary people can use this tool to create videos. Users only need to upload a 2-minute short video to create a digital human like a real person, and even "subtle body language" such as gestures, facial features, and mouth movements can be adjusted.
Joshua Xu personally stepped forward and launched a digital character that aligns with her own image.
At present, its customers include NVIDIA, Amazon, Volvo, and others. The products support over 300 voices, over 40 languages, and provide over 100+digital human materials and templates. It has a complete range of advertising, e-commerce, news, and more, and can also edit and generate PPTs on its own.
According to Tianyancha data, Shiyun Technology has received a cumulative investment of nearly ten million yuan from venture capitalists such as Sequoia China and Zhenge Fund.
It is worth noting that HeyGen is still a profitable AI application. This tool has developed a very detailed payment plan, categorizing the personal/enterprise versions, with options for monthly or annual packages. It provides plugins for premium use, and for ordinary users, the annual package price is evenly distributed, costing $24 per month.
In a blog post released at the end of April this year, the company revealed that after the product was launched in July 2022, after 178 days, the company's annual recurring revenue (ARR) had reached $1 million, reaching a profitable state after 7 months of launch.
Joshua Xu once stated that HeyGen has accumulated millions of video data and will also consider building a Large Model (LLM) in the vertical field of video.