Today in Generative Media
Future of AI in book publishing; Etsy's genAI seller guidelines; AI interior design apps
News
Is AI the Bitter End—or the Lucrative Future—of Book Publishing? (Esquire)
Etsy adds AI-generated item guidelines in new seller policy (TechCrunch)
AI design apps made my new apartment look odd (The Verge)
Power-hungry AI is driving a surge in tech giant carbon emissions. Nobody knows what to do about it (The Conversation)
Research
MVSGaussian Fast Generalizable Gaussian Splatting Reconstruction from Multi-View Stereo (ECCV 2024, project page)
3D Gaussian Ray Tracing: Fast Tracing of Particle Scenes (project page)
4DiM: Controlling Space and Time with Diffusion Models (project page)
Video-to-audio generation with hidden alignment (project page)
The Unmet Promise of Synthetic Training Images: Using Retrieved Real Images Performs Better (arXiv)
VEnhancer: Generative Space-Time Enhancement for Video Generation (arXiv)
Neural Localizer Fields for Continuous 3D Human Pose and Shape Estimation (arXiv)
PaliGemma: A versatile 3B VLM for transfer (arXiv) PaliGemma report will hit arxiv tonight. (X)
TL;DR: it's all about the loss surfaces (X) [a long, technical post about training GANs more robustly]

