Today in Generative Media

Meta's genAI selfies; Hollywood law firm fights celebrity deepfakes; Stable Video 4D

News

• Meta Launches AI Selfie Generator ‘Imagine Me’ (PetaPixel)

• Top Hollywood Law Firm Launches Service to Combat Soaring Celebrity Deepfake Phenomenon (Variety)

• Stability AI steps into a new gen AI dimension with Stable Video 4D (VentureBeat)

• If A.I. Is Coming for Comedy Writers, Simon Rich Is Ready (New York Times)

Software

• AI arms race escalates: OpenAI offers free GPT-4o Mini fine-tuning to counter Meta’s Llama 3.1 release (VentureBeat)

• Mistral shocks with new open model Mistral Large 2, taking on Llama 3.1 (VentureBeat)

• End Frame is now available in Luma DreamMachine! (X)

• Open-Sora Plan v1.2.0 is here! (X) [An open source project attempting to reproduce what’s publicly known about Sora.]

• Managed to get LivePortait running near realtime within ComfyUI, now also without Insightface, using Google's MediaPipe (Apache-2.0 license) instead! (X)

• PyTorch 2.4.0 comes out today (X)

Research

• VFusion3D: Learning Scalable 3D Generative Models from Video Diffusion Models (ECCV 2024, project page)

• SV4D: Dynamic 3D Content Generation with Multi-Frame and Multi-View Consistency (project page)

• MINT-1T: Scaling Open-Source Multimodal Data by 10x: A Multimodal Dataset with One Trillion Tokens (project page)

• IMAGDressing-v1 : Customizable Virtual Dressing (project page)