Respeecher Secures New Funding with an Ethics-First Approach to AI Voice Cloning

Ukrainian Synthetic Voice Startup Respeecher Thrives Amid Challenges

Ukrainian synthetic voice startup Respeecher is achieving remarkable success despite the challenging conditions in their city caused by ongoing conflict and significant competition in the voice synthesis market. Their recent $1 million funding round will enable the company to expand its services to additional studios and enhance its offerings for media and gaming clients.

Respeecher gained prominence for its groundbreaking work in replicating iconic voices, famously recreating James Earl Jones's portrayal of Darth Vader for a Star Wars animated series, and providing a younger voice for Luke Skywalker in The Mandalorian. The company has also collaborated with renowned game developer CD Projekt, known for the Witcher and Cyberpunk franchises, and recently partnered with Warner Music to recreate the voice of legendary French singer Edith Piaf.

Unlike conventional text-to-speech engines, Respeecher’s innovative approach employs voice models to alter the speech of actors, who work diligently to embody the desired voice or character. This method creates a more authentic experience, akin to a prosthetic voice. Respeecher also offers accent modification, which can help eliminate unwanted accents or enable actors to adopt different ones.

The ethical implications of cloning voices, particularly those of long-deceased individuals who cannot consent, are significant. Some startups may overlook these concerns, limiting their client base. In contrast, Respeecher prioritizes ethics across all aspects of its business.

“Consent is obtained from rights holders; for deceased actors, this often involves their estate or family,” explained CEO and co-founder Alex Serdiuk. “Many times, these stakeholders are directly involved in the process, offering invaluable feedback to ensure the voice is perfect. Such projects serve as tributes to their loved ones and the characters they created.”

Respeecher recently partnered with Calm to develop a synthetic voice modeled after the iconic actor Jimmy Stewart.

For living actors, permission and compensation arrangements are established from the outset. As the industry evolves, voice actors are beginning to view these voice models as valuable assets to be controlled and monetized rather than as threats to their careers. Respeecher is curating a library of voices from actors who have opted into this innovative process and has also joined Adobe’s Content Authenticity Initiative.

While Respeecher may have forgone rapid scaling during this pivotal year in AI, their deliberate approach might ultimately prove beneficial. After all, the team has been navigating significant challenges within Kyiv over the past year.

“Like all Ukrainian businesses and startups, this war has redefined our understanding of resilience,” said Serdiuk. “Securing funding is always a challenge, and it would be easier without the constant threat of missile strikes or drone attacks. After experiencing this, I believe there are few challenges our team cannot overcome or solutions we cannot find.”

Amidst these trials, the company has ventured into a new domain: creating synthetic voices for individuals who can no longer speak independently. While this area may not be as lucrative, it holds the potential to transform lives.

For example, Respeecher has teamed up with hospitals and patients affected by conditions such as ataxia and laryngectomy. One of their notable projects involved working with Konrad Zieliński, a PhD student at the University of Warsaw who lost his voice due to laryngectomy. "Our technology enabled him to communicate more naturally in his own voice,” Serdiuk noted. Further details about Konrad's journey can be found in a dedicated blog post.

Today, Respeecher announced its successful completion of a $1 million pre-Series A funding round, with contributions from entrepreneur Gary Vaynerchuk and investment funds ffVC, Bad Ideas, ICU, and SID Venture Partners.

Keywords: AI, Media, Entertainment, Respeecher, Speech Synthesis, Synthetic Speech