Source: Caijing News Author: Shi Zhengcheng
Synthesia's last financing was in June of last year, reaching a valuation threshold of $1 billion after investment. In addition to more realistic AI 'Avatar,' this update also greatly reduces the difficulty of video production, and cross-language video translation is now 'one-click direct.'
On Monday, local time, Synthesia, an AI startup based in London, UK, announced the launch of a series of updates to its "virtual humans" generation products, aiming to push the current business society from the "text world" to a "video-first world".
(Source: Synthesia official website)
In general, the updates from Synthesia include "half-body" AI avatar (now with gestures!), AI video production assistant, AI screen recording function, and AI dynamic (translation) player.
According to public market data, the last financing for Synthesia was in June of last year, led by California venture capital fund Accel and strategic investment from Nvidia. Including the last round of financing of $90 million, Synthesia's total financing scale reached $156 million, and the post-investment valuation reached the threshold of $1 billion.
The company's CEO, Victor Riparbelli, told the media at the time that the company already had 50,000 customers and did not need additional financing, but Accel and Nvidia came to offer an attractive quote.
From "AI avatar" to "AI half-body avatar"
As Synthesia's flagship product, personalized AI avatar is still the product that the company focuses on. There are two major changes in this update. The process of making AI avatars is easier, and the next stage of AI avatars will be half-body avatars-making body posture and gestures part of AI "Avatar".
According to the company, users can use a network camera or mobile phone to create custom avatars. At the same time, this avatar can synchronize with natural lip movements and voice, replicating the user's voice in more than 30 different languages. An English-only speaker can use an AI avatar to "tell" German, Spanish, Chinese, and other languages to different people.
Previously, to make a professional "AI avatar", users had to personally enter the studio for data collection.
Another major upgrade of the new product is that AI "Avatar" is no longer just a head. Body language and gestures can also be generated through AI.
Redefining the video production process
Another important update from Synthesia is an AI video assistant, which greatly simplifies the process from style layout, video editing to cross-language translation, using AI.
Earlier this year, Synthesia solved the problem of using AI to write video scripts. Users can upload background materials, specify tone, video length, audience, and other parameters, and then let AI complete the writing. In subsequent updates, AI video assistants will incorporate brand identities (layout, color, logos, etc.) to keep the look and feel of these types of videos consistent.
According to the company, in conversations with customers this year, someone suggested converting "hundreds" of explanatory documents into videos, which would help customers find answers more easily and save customer service teams resources. Therefore, the next AI video assistant will provide batch video creation services and one-click conversion of articles into high-quality videos by setting templates.
Considering that some business users have screen recording needs, Synthesia has also integrated a tool in the system to capture screens, record, transcribe subtitles, and match the timeline. If necessary, users can also insert "speaking" personal avatars into the video.
Finally, there is a player that automatically aligns the translated content. Last year, Synthesia launched a video translation feature that could translate video content into more than 120 languages. But this product also has a drawback: once the user needs to adjust the video, all the translations have to be managed one by one. In this update, users only need to adjust and edit the original video, and the other translated videos will be updated automatically.
In this update, users only need to adjust and edit the original video, and the other translated videos will be automatically updated.
It is reported that Synthesia will also launch a video player next month that can translate the entire video into the viewer's language with just one click.
Editor/Lambor