Today in Generative Media

Adobe Firefly's safe GenAI video; Meta the AI behemoth; Beyond Presence gets $3.1M

News

• Adobe Firefly Video Is the First Commercially Safe Generative AI Video Model (PetaPixel)

• How Meta has become an AI behemoth (Yahoo Finance)

• After selling his last AI startup to Meta, Beyond Presence’s founder nabs $3.1M to build lifelike avatars (TechCrunch)

• Duolingo CEO Luis von Ahn wants you addicted to learning (The Verge)

• OpenAI's AI-adjusted earnings numbers have echoes of Groupon and WeWork (Business Insider)

• AI Is Definitely Coming for Your Job (CNET)

Software

• Google’s Powerful AI Image Generator is Now Available on Gemini (PetaPixel)

• Stunning! The AI Model DIAMOND Can Simulate Counter-Strike: Global Offensive on a Single RTX 3090 (AIBase)

  • Ever wanted to play Counter-Strike in a neural network? (X)

  • Run AI Generated World on Your Laptop (X)

• Introducing ComfyDeploy v2. The operating system for Gen AI (X)

• Flux.1 Dev - ControlNet Inpainting - Beta now on tost.ai , runpod.io and ComfyUI (X)

Research

• Can Visual Foundation Models Achieve Long-term Point Tracking? (ECCV 2024, project page)

• MimicTalk: Mimicking a personalized and expressive 3D talking face in few minutes (NeurIPS 2024, project page)

• TANGO: Co-Speech Gesture Video Reenactment with Hierarchical Audio-Motion Embedding and Diffusion Interpolation (project page)

• F5-TTS: A Fairytaler that Fakes Fluent and Faithful Speech with Flow Matching (project page)

• Representation Alignment for Generation: Training Diffusion Transformers Is Easier Than You Think (project page)

• Loong: Generating Minute-level Long Videos with Autoregressive Language Models (project page)

• ControlAR: Controllable Image Generation with Autoregressive Models (arXiv)

• DisEnvisioner: Disentangled and Enriched Visual Prompt for Customized Image Generation (project page)

• Conditional Image Synthesis with Diffusion Models: A Survey (arXiv)

• MeshGS: Adaptive Mesh-Aligned Gaussian Splatting for High-Quality Rendering (arXiv)

Misc

• Made this video with iPhone photos I took of my friend Stephanie that I used as keyframes in LumaLabsAI! (X)

• You really, really should not trust audio clips anymore (X)

• Crazy how AI with personalised model now works as brain extension with locked-in aesthetics (X)