Today in Generative Media

Canva vs. Adobe with AI; Google's 'Reimagined' wrecks; Hotshot's text-to-video genAI

News

• Canva Steps Up To Challenge Adobe With AI Acquisition And Partnerships (Forbes)

• Google’s AI ‘Reimagine’ tool helped us add wrecks, disasters, and corpses to our photos (The Verge)

• Hotshot launches new text-to-video AI generator (VentureBeat)

• Symphonic Opens Up Catalog to Train AI Models Through Musical AI Partnership (Billboard)

Software

• Try SpaRP to reconstruct 3D textured mesh from one or a few unposed images! (HuggingFace)

• This is the repo for the paper Magic Fixup: Streamlining Photo Editing by Watching Dynamic Videos (GitHub)

Research

• DEGAS: Detailed Expressions on Full-Body Gaussian Avatars (project page)

• Robust 3D Gaussian Splatting for Novel View Synthesis in Presence of Distractors (project page)

• MeshFormer: High-Quality Mesh Generation with 3D-Guided Reconstruction Model (project page)

• MegaFusion: Extend Diffusion Models towards Higher-resolution Image Generation without Further Tuning (project page)

• ShapeSplat: A Large-scale Dataset of Gaussian Splats and Their Self-Supervised Pretraining (project page)

• Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Model (arXiv)