Alibaba запустив emo модель яка оживляє фото для відео з говором або співом

🎙🎙 Introducing Alibaba's EMO — an Audio2Video model that breathes life into photographs, allowing them to speak or sing

🤖 The AI innovation, EMO (Emote Portrait Alive), crafts highly realistic videos featuring "talking heads" from a single image and an audio file.

🪄 Capabilities of EMO

This model excels in syncing lip and head movements in photos with an accompanying audio track, achieving a remarkably lifelike effect. The result is a video where the subject not only "sings along" but also exhibits a range of facial expressions, emotions, and gestures.

🎶 From Mona Lisa to Eminem

The technology was showcased using iconic figures. For instance, the Mona Lisa was made to recite a passage from Shakespeare, while a photo of Leonardo DiCaprio was animated to perform an Eminem track. Particularly astounding was the demonstration where Joaquin Phoenix, as the Joker, was given the voice of Heath Ledger.

ℹ️ Alibaba Group, a Chinese tech giant active in e-commerce and cloud computing, is now fervently investing in generative AI research and development. In October 2023, at its annual Hangzhou conference, the company unveiled the latest iteration of its AI model — Tongyi Qianwen 2.0, powered by hundreds of billions of parameters.

We're looking forward to when EMO becomes available to everyone 🙂

#news @hiaimediaen