Today in Generative Media

Meta's AI glasses: Real-time doxxing and no training opt-out; OpenAI raises $6.6B

News

• College students used Meta’s smart glasses to dox people in real time (The Verge)

  • I-XRAY: The AI Glasses That Reveal Anyone’s Personal Details—Home Address, Name, Phone Number, and More—Just from Looking at Them (project page)

• Meta confirms it may train its AI on any image you ask Ray-Ban Meta AI to analyze (TechCrunch)

• OpenAI just raised $6.6 billion to build ever-larger AI models (The Verge)

Software

• Announcing FLUX1.1 [pro] and the BFL API (Black Forest Labs)

• Introducing Invoke 5.0 – Our Biggest Update Yet.

  • Invoke - Professional Creative AI Tools for Visual Media (GitHub)

  • You can get it on http://pinokio.computer (X)

• Voice Restoration with Transformer-based Model: Upload a degraded audio file or select an example, and the space will restore it using the VoiceRestore model (HuggingFace)

• CoTracker2.1 is now available! (GitHub)

• StableDelight: Revealing Hidden Textures by Removing Specular Reflections (GitHub)

• GraphRAG-UI is a user-friendly interface for GraphRAG, a powerful tool that uses the Retrieval-Augmented Generation (RAG) approach to index and query large text data. (GitHub)

Research

• StableNormal: Reducing Diffusion Variance for Stable and Sharp Normal (SIGGRAPH Asia 2024, project page) Demo on HuggingFace. Code on GitHub.

• Exact Volumetric Ellipsoid Rendering for Real-time View Synthesis (project page)

• GaussianBlock: Building Part-Aware Compositional and Editable 3D Scene by Primitives and Gaussians (arXiv)

Misc

• Shadow puppets are probably the purest/simplest live performance you can do - reimagined with these amazing tools! (X)