Today in Generative Media

No politics in Canva AI; Developers vs. GitHub Copilot; OpenAI blocks China access

News

• You can't use Canva's AI to create political flyers or spread medical misinformation, the CEO said (Business Insider)

• The developers suing over GitHub Copilot got dealt a major blow in court (The Verge)

• Chinese developers scramble as OpenAI blocks access in China (The Guardian)

Software

• Pestle’s app can now save recipes from Reels using on-device AI (TechCrunch)

• Adobe adds CAI 'Content Credentials' to Camera Raw, Lightroom and Photoshop (DPReview)

• Galaxy AI may soon power real-time translation in WhatsApp, and maybe Google Meet too (9to5google.com)

• We're thrilled to tease Diva by Synthetique. 📸✨ Something revolutionary is coming to product and fashion photography — set to redefine in-studio creativity by combining: Automated Photography x Virtual Production x Generative AI x Robots (LinkedIn)

• PaintsUndo: A Base Model of Drawing Behaviors in Digital Paintings (project page) Paints-Undo presents a family of models that take an image as input and then output the drawing sequence of that image. (GitHub)

Research

• Interactive 3D Scene Generation from a Single Image (project page)

• TAPVid-3D: A Benchmark for Tracking Any Point in 3D (project page)

• UltraEdit UltraEdit: Instruction-based Fine-Grained Image Editing at Scale (project page)

• Lifelike Audio-Driven Portrait Animations through Editable Landmark Conditioning (project page)