Nvidia-Backed Synthesia Introduces ‘Expressive Avatars’ With Human Emotions
Video screenshot via Synthesia
Gone are the stiff mannequins of yesterday’s chatbots—Synthesia’s new artificial intelligence avatars have shed their poker faces and are learning to emote. Backed by tech giant Nvidia, Synthesia, an AI firm focused on avatar creation, has unveiled a new generation of avatars dubbed Expressive Avatars that can convey a whole spectrum of human emotions based on what you tell them.
This new army of characters can display emotions such as happiness, sadness, and frustration, offering a more nuanced and realistic interaction model for users.
Synthesia’s technology operates by analyzing extensive video data collected from human actors. The actors perform various emotional states in front of a green screen, which the AI system then uses to learn the nuanced facial expressions and body language associated with each emotion. This data enables the creation of avatars that respond emotionally to text inputs provided by users, making digital interactions feel more authentic and human-like.
Based in London, Synthesia has recently attracted significant investment, raising US$90 million last year and reaching a valuation close to US$1 billion. This influx of capital underscores the potential impact and commercial viability of their technology. Currently, over 55,000 businesses, including many from the Fortune 100, utilize Synthesia’s avatars for tasks ranging from corporate presentations to training videos. By eliminating the need for traditional video production elements such as cameras and on-screen talent, Synthesia claims to significantly cut the costs and complexity of content creation.
Video screenshot via Synthesia
The applications for Expressive Avatars are vast. In customer service, AI avatars could not only provide information but also relate empathetically to customers’ concerns. Educational tools could become more engaging, utilizing avatars to simulate teacher-student interactions. Moreover, the entertainment sector might see new storytelling techniques emerge, with avatars enabling novel forms of narrative expression.
Video screenshot via Synthesia
However, the advancement also brings challenges, particularly in the potential for misuse in creating deceptive or misleading content, such as fake news. To mitigate these risks, Synthesia insists that those wishing to create content with its avatars become enterprise customers. It has also established a robust framework to ensure its technology is used ethically, safeguarding against abuse.
[via Bloomberg, NBC Washington, CNBC, MIT Technology Review, videos and screenshots via Synthesia]
