Today in Generative Media

OpenAI's Sora finally launches; GenAI revolutionizes news; Variety's genAI report

News and Opinion

• Sora is here (OpenAI)

  • OpenAI’s video generator, Sora, aims to kickstart the AI video era (Washington Post)

  • OpenAI’s Sora is launching today — here are highlights from the first review (TechCrunch)

  • OpenAI has finally released Sora (The Verge)

• Generative AI will revolutionize the news industry (UNC Hussman School of Journalism)

• Generative AI, celebrity deepfakes & digital replicas: A special report (Variety)

• Innovation meets litigation: How media companies are tackling AI’s complex impact (Digiday)

• Hinge CEO: Searching for a partner in real life won’t ‘hold a candle to’ meeting online with AI (semafor.com)

• The phony comforts of AI skepticism (Platformer)

• AI and Creativity: A Backstage Perspective (Jaclyn Konzelmann on Substack)

• Four loose ends (job loss, art-washing, & more) (Mike Gioia on Substack)

Software

• BRIA Generative Fill API (HuggingFace)

• X gives Grok a new photorealistic AI image generator (The Verge)

• Hunyuan Video model is now 4x to 5x faster on FAL without compromising any quality! We are also reducing the price to $0.4 per video! (X)

Research

• Generative Photography: Scene-Consistent Camera Control for Realistic Text-to-Image Synthesis (project page)

• Negative Token Merging: Image-based Adversarial Feature Guidance (project page)

• InstantSwap: Fast Customized Concept Swapping across Sharp Shape Differences (project page)

• ControlFace: Harnessing Facial Parametric Control for Face Rigging (project page)

• MEMO: Memory-Guided Diffusion for Expressive Talking Video Generation (project page)

• FLOAT: Generative Motion Latent Flow Matching for Audio-driven Talking Portrait (project page)

• Synergizing Motion and Appearance: Multi-Scale Compensatory Codebooks for Talking Head Video Generation (project page)

• VISION XL: High Definition Video Inverse Problem Solver using Latent Diffusion Models (project page)

• Sharp-It: A Multi-view to Multi-view Diffusion Model for 3D Synthesis and Manipulation (project page)

• Neural Global Illumination via Superposed Deformable Feature Fields (SIGGRAPH Asia 2024, ACM Digital Library)

• Fitting Spherical Gaussians to Dynamic HDRI Sequences (SIGGRAPH Asia 2024, ACM Digital Library)

• Novel View Synthesis with Pixel-Space Diffusion Models (arXiv)