This weekend in Generative Media
OpenAI/Google AI war; Music biz' Anthropic suit wobbles; Apple's false AI headlines
[I will be on vacation and/or traveling the next several weeks. Expect fewer editions between today and early January!]
News and Opinion
The AI war between Google and OpenAI has never been more heated (Ars Technica)
Federal Judge Reportedly Prepares to Dismiss the Majority of Music Publishers’ Anthropic Lawsuit (Digital Music News)
Apple called on to ditch AI headline summaries after BBC debacle (The Register)
TCL’s new AI short films range from bad comedy to existential horror (The Verge)
Arizona School’s Curriculum Will Be Taught by AI, No Teachers (Gizmodo)
Here’s the full list of 49 US AI startups that have raised $100M or more in 2024 (TechCrunch)
Software
HunyuanVideo Native Support in ComfyUI (ComfyUI blog)
Introducing Meta Video Seal: A state-of-the-art, open-source model for video watermarking (Meta AI)
Arbitrary-steps Image Super-resolution via Diffusion Inversion (GitHub)
Research
Wonderland: Navigating 3D Scenes from a Single Image (project page)
InstantRestore: Single-Step Personalized Face Restoration with Shared-Image Attention (project page)
ReF-LDM: A Latent Diffusion Model for Reference-based Face Image Restoration (project page)
SharpDepth: Sharpening Metric Depth Predictions Using Diffusion Distillation (project page)
IDArb: Intrinsic Decomposition for Arbitrary Number of Input Views and Illuminations (project page)
Enhance-A-Video: Better Generated Video for Free (project page)
MeshArt: Generating Articulated Meshes with Structure-guided Transformers (project page)
SimAvatar: Simulation-Ready Avatars with Layered Hair and Clothing (project page)
BLADE: Single-view Body Mesh Learning through Accurate Depth Estimation (project page)
MCMat: Multiview-Consistent and Physically Accurate PBR Material Generation (project page)
Motion-2-to-3: Leveraging 2D Motion Data to Boost 3D Motion Generation (project page)
BrushEdit: All-In-One Image Inpainting and Editing (project page)
PrEditor3D: Fast and Precise 3D Shape Editing (project page)
ColorFlow: Retrieval-Augmented Image Sequence Colorization (project page)
Misc
How to fine-tune open LLMs in 2025 with Hugging Face (Philipp Schmid’s blog)