Synthesia raises $50 million to create synthetic videos with AI - Actu IA

Synthesia is a startup using AI to create synthetic avatar videos for marketing purposes. After an initial round of funding in April 2021, it announced a new $5O million round led by Kleiner Perkins last December, with participation from GV and existing investors Firstmark Capital, LDV Capital, Seedcamp and MMC Ventures. Synthesia was founded in 2017 by a team of researchers and entrepreneurs from UCL, Stanford, TUM and Cambridge who want to replace cameras with code, make everyone a creator and for synthetic media to progress in an impactful and ethical way." Synthetic media refers to video, image, text and voice that have been wholly or partially generated by computers. The ability of AI-driven systems to generate audiovisual content is a development made possible by the latest advances in deep learning. Synthesia uses deep learning to create visual chatbots, product videos, and sales videos for clients without actors, camera crews, studios, or cameras. A video production is expensive, requires studios, actors, cameras, post-production can be complex. Synthesia offers the realization of videos in a few minutes from a text, integrated avatars if the customer has not created one himself, he chooses a real or synthetic voice and the video is ready, with a possible translation in 50 languages. Afterwards, it is possible to modify the content of the video, to add text, images… Since the fundraising in April, Synthesia has added features that allow users to create their own animated talkers even more easily, and the platform now has 1,000 custom avatars in use. Riparbelli cited Ernst & Young as an example of a customer. The company has 35 partners with their own avatars, creating videos for internal and customer communications. Companies looking to make video creation easier with AI and avatars still need to increase realism, but they also need to ensure user safety and the credibility of their own platforms. An Israeli company called D-ID demonstrated its technology at Disrupt 2021, which can take a still image of a person and turn it into video content. "We will not offer our software for public use.