This weekend in Generative Media

Fake movie trailers vs. AI slop; Trump's AI selfie as Pope; Zuck's banal AI vision

News and Opinion

• Fake movie trailers were an art form. Then came the AI slop. (Washington Post)

• Trump Posts AI-Generated Image of Himself Dressed as the Pope (Variety)

• Mark Zuckerberg’s Banal AI Vision (New York Intelligencer)

• “The article will die, should die, but storytelling will not”: Notes from the Nordic AI in Media Summit (Nieman Lab)

• DOGE Is in Its AI Era (Wired)

• Google Plans to Roll Out Its A.I. Chatbot to Children Under 13 (New York Times)

Software

• FlexiCubes is now under Apache 2.0! (X)

• [Hedra] Character-3 API is now available! (X)

• This repository contains the example inference script for the MagicPortrait-preview model. (GitHub)

Research

• Eye2Eye: A simple approach for monocular-to-stereo video synthesis (project page)

• LaRI: Layered Ray Intersections for Single-view 3D Geometric Reasoning (project page)

• RePerformer: Immersive Human-centric Volumetric Videos from Playback to Photoreal Reperformance (project page)

• RayZer A Self-supervised Large View Synthesis Model (project page)

• JointDiT: Enhancing RGB-Depth Joint Modeling with Diffusion Transformers (project page)

• Real-Time Animatable 2DGS-Avatars with Detail Enhancement from Monocular Videos (arXiv)

• Controllable Weather Synthesis and Removal with Video Diffusion Models (arXiv)

• FreBIS: Frequency-Based Stratification for Neural Implicit Surface Representations (arXiv)

• Can We Achieve Efficient Diffusion without Self-Attention? Distilling Self-Attention into Convolutions (arXiv)

Misc

• Star Wars Changed Visual Effects — AI Is Doing It Again | Rob Bredow | TED (YouTube)