Alejandro Franceschi’s Post

#ByteDance showcases #realtime fully #genaivideo, with #avatars of all types. What's freakish is that these execute interactive *non-verbal* cues, as they listen and respond. That is the kind of artful subtlety that can *manipulate people* with ever more convincing #genai. I've presented a company that does this via an API with super low latency using a kind of #deepfake, but this version by ByteDnce (which owns #TikTok), can use content that is fully synthetic, or an alteration of a real person. This is happening simultaneously in #volumetric applications, so this can exist in #videogames, #vfx, #animation, and #xr. Abstract: INFP: Audio-Driven Interactive Head Generation in Dyadic Conversations (Dyadic = social interactions between two persons) We present INFP, an audio-driven interactive head generation framework for dyadic conversations. Given the dual-track audio in dyadic conversations and a single portrait image of arbitrary agent, our framework can dynamically synthesize verbal, non-verbal and interactive agent videos with lifelike facial expressions and rhythmic head pose movements. Additionally, our framework is lightweight yet powerful, making it practical in instant communication scenarios such as the video conferencing. INFP denotes our method is Interactive, Natural, Flash and Person-generic. I must add that without the availability of the code, it's difficult to emulate. Given what I have seen from other companies, from cloud LCM models for 2D, 3D, lip-sync, and the plethora of gen ai video updates from the likes of #OpenAI and #Sora to #Google's #Veo2 ("veo" means "I see" in Spanish), this seems like a naturally progressive jump. Project Page: https://lnkd.in/gjZdpC5f *** "The Party told you to reject the evidence of your eyes and ears. It was their final, most essential command." - George Orwell, "1984"

Yalın Solmaz

GenAI Advisor to Creatives 🤖 | Minimax, Nim Creative Partner 🌈 | Thought Leader & Speaker 🎤 | Creator Economy Veteran 🎥 | Ex-YouTube

14h

They’re either developing this tech to build virtual influencers or to replace humans stuck in endless corporate zoom meetings. Joke aside, we’re getting to the point where we cross the uncanny valley here. Once beyond it, we won’t be able to believe anything we see on social media.

Luka Tisler

CEO | Founder - 6 Fingers

15h

Probably another code that won't see the light of day.

See more comments

To view or add a comment, sign in

Explore topics