Synthesia, a synthetic media startup, has partnered with generative AI voice platform Celebrities to power its deepfake video host with voice clones and AI-generated audio. Synthesia uses generative AI to generate people's appearances to realistically mimic them, but Eleven Labs' artificial audio can enhance that even further.
Synthesia Eleven Lab
Synthesia's platform employs AI to analyze real people in videos, allowing users to fine-tune background details and other elements to generate simulacra or completely fictional people acting out from a script. . This includes voice and other sounds, but this partnership will see Synthesia integrate Eleven Labs' more advanced voice generation model for businesses on the Enterprise plan. . Synthesia users will now be able to create high-quality, natural-sounding vocals for their video projects. This process will be faster and more efficient than the standard option.
“This collaboration is a step forward in our ongoing commitment to adding even more innovative capabilities to our Generative AI Video Platform, helping the UK technology company pioneer the adoption of responsible Generative AI and helping drive the adoption of responsible Generative AI across the world. ” explained Guillaume Boniface-Chang, Head of Product at Synthesia, in a blog post. “Customized voices for his avatars, including clones, can also speak multiple languages. This is a requested feature.
Boniface-Chang said the collaboration also provides users with the opportunity to enhance their custom avatars to perform in multiple languages. The idea is to help companies looking to localize their content for international audiences, a frequent request from Synthesia's clients.
Synthesia also highlighted how the company will address the growing potential for deepfakes to impersonate real people with less-than-benevolent intentions. Customers considering creating custom avatars or using voice clones must follow Synthesia's rules regarding consent and control. This means that avatars cannot be created without the explicit permission of the individual whose likeness is used, and users can control how their avatar is used. Furthermore, Synthesia undertakes to address any request to remove the user's data and likeness from its database.
It's not just a theoretical concern. Eleven Lab's technology is widely in demand due to its faithfulness to real voices, and was used by former Pakistani Prime Minister Imran Khan in his speech from prison. It was adopted. In addition, robocalls to New Hampshire voters earlier this year used ElevenLabs to suppress voter turnout in the state's primary election. created a deepfake version of , which violates Eleven Labs' own rules. An investigation revealed that the calls were traced to Life's Lingo, the communications provider that sent the calls on behalf of his corporation. The FCC issued a cease-and-desist order and subsequently banned deepfake robocalls completely. This is why the development of deepfake detectors by companies such as Pindrop and internal detectors such as Eleven Labs is similarly fast-tracked.
Follow @voicebotaiFollow @errichschwartz
Synthesia, a synthetic media AI startup, raises $90 million
Eleven Labs raises $80M to share generative AI voice models, tools and deepfake voice marketplace
Pindrop launches real-time audio deepfake detection tool Pindrop Pulse
