The startup behind the popular MyHeritage app, which turned traditional family photos into lifelike moving portraits, is returning with a new application of its technology: transforming still photos into ultra-realistic video that can say whatever you want.
D-Speaking ID’s Portraits may resemble the infamous “deepfakes” that have made headlines in recent years, but the underlying technology is entirely different, and basic functionality requires no training.
D-ID, which launched with a totally different focus (scrambling facial recognition software) at TechCrunch Battlefield in 2018, introduced its new Speaking Portraits product live at TechCrunch Disrupt 2021. The company demonstrated a variety of applications, including using its new technology to create a multilingual TV anchor capable of expressing various emotions, virtual chatbot personas for customer support interactions, professional development training courses, and interactive conversational video ad kiosks.
This new product, as well as D-relationship ID’s with MyHeritage, which saw the latter’s app momentarily take over the top of Apple’s App Store charts, are clearly significant diversions from the company’s primary focus. D-ID was still raising funds using its previous strategy until May of last year, when it announced a partnership with MyHeritage in February, followed by a similar deal with GoodTrust in March, and a big tie-in with Warner Bros. on the Hugh Jackman film “Reminiscence” that allowed fans to insert themselves into the trailer.
D-turnaround ID’s may appear more dramatic than others, but its new focus on bringing photographs to life isn’t that far from its de-identification software from a technical standpoint. According to D-ID CEO and co-founder Gil Perry, the firm changed course since it was clear that there was a significant addressable market for this type of application. Clients like Warner Bros. and an App Store-dominating app from a relatively obscure brand seem to back up that claim. Speaking Portraits, on the other hand, is aimed at both large and small businesses, and allows anyone to create a full HD movie from a source image and recorded speech or typed text.
D-ID is starting the product with support for English, Spanish, and Japanese, but it wants to expand to more languages when users request it. D-ID offers two types of Speaking Portraits, one of which is a “Single Portrait” that can be created from a single still image and includes an animated head but no other moving components. This one will only work with the photo’s existing background.