Alibaba Cloud’s Emote Portrait Alive (EMO) model,Watch The Temptation of Kimono (2009) which is able to convert portrait photos into talking and even singing avatars is now available on its Tongyi Qianwen app, in a move the tech giant hopes will bring a wider user base to test its GenAI tool. The tool doesn’t currently support custom audio, but users can choose from more than 80 templates, including hit songs, and generate videos after uploading a portrait photo. In a WeChat post on Alibaba Cloud’s account, the company said EMO had become “one of the most popular AI models after OpenAI’s Sora” after it made EMO-related papers public in late February. The cloud computing unit of Alibaba also added that its audio-driven technology is expected to be applied to fields such as digital humans, movie production, and virtual accompaniment in the future. [Alibaba Cloud, in Chinese]
