Cogniscendo

Cogniscendo

Uncanny Valley Recession

AI Waifus and Husbandos are on the way

Prakash's avatar
Prakash
Feb 29, 2024
∙ Paid
🔷 Subscribe to get breakdowns of the most important developments in AI in your inbox every morning.

Here’s today at a glance:

  • The Pictures Learn To Talk

  • Things happen

  • AI artwork of the day

📷 The Pictures Learn To Talk

The AI team at Alibaba came up with a milestone innovation: generating realistic talking people from just a single portrait image and audio. Some samples:

And the full paper via video:

Let’s break down what they’ve done:

  • You send the model a single portrait image of a face an audio file of a single person talking, speaking, or singing

  • The model generates a video of the face on the image talking, speaking, or singing near perfect lipsync in multiple languages true emotion, eyes, eyebrows, jaw, cheeks, and facial angles including natural face movement, hair follow-through

It’s an amazing, state-of-the-art piece of technology that will change the world the moment it’s fully released. The implications for YouTube, TikTok, Instagram, customer service videos, and the…

User's avatar

Continue reading this post for free, courtesy of Prakash.

Or purchase a paid subscription.
© 2026 8teapi · Privacy ∙ Terms ∙ Collection notice
Start your SubstackGet the app
Substack is the home for great culture