This weekend in Generative Media

Microsoft, Perplexity AI on content theft; The deluge of bonkers AI art; More lawsuits...

News

• Microsoft’s AI boss thinks it’s perfectly okay to steal content if it’s on the open web (The Verge)

• Perplexity’s grand theft AI (The Verge)

• The deluge of bonkers AI art is literally surreal (Washington Post)

• The Center for Investigative Reporting is suing OpenAI and Microsoft (The Verge)

• An AI-designed horse purse is tearing apart this small but passionate community (The Verge)

• Here comes a Meta Ray-Bans challenger with ChatGPT-4o and a camera (The Verge)

Software

• I Turned 15 Celebrities Into Shrek Because WHY NOT — You Can Use Our Generator To Make 'Em Too (BuzzFeed)

• The AI voice generator for video creators Find your voice with the ultimate text-to-speech generator, featuring top voice actors exclusive to Artlist. (artlist.io)

• Automatic video editing Turn any text, article or story into a video, in seconds (tellers.ai)

• Run Gemma 2 9B in a free Google Colab! The best LLM < 15B all in less than 5 lines of code! Enjoy! (X)

• Disentangled Motion Modeling for Video Frame Interpolation (GitHub)

Research

• TexPainter: Generative Mesh Texturing with Multi-view Consistency (SIGGRAPH 2024, project page)

• A Diffusion Approach to Radiance Field Relighting using Multi-Illumination Synthesis (EGSR 2024, project page)

• MIRReS: Multi-bounce Inverse Rendering using Reservoir Sampling (project page)

• Fast and Uncertainty-Aware SVBRDF Recovery from Multi-View Capture using Frequency Domain Analysis (project page)

• Portrait3D: 3D Head Generation from Single In-the-wild Portrait Image (project page)

• ResMaster: Mastering High-Resolution Image Generation via Structural and Fine-Grained Guidance (project page)

• Dynamic Gaussian Marbles for Novel View Synthesis of Casual Monocular Videos (arXiv)

• Neural Assets: 3D-Aware Multi-Object Scene Synthesis with Image Diffusion Models (project page)

• GS-ROR: 3D Gaussian Splatting for Reflective Object Relighting via SDF Priors (arXiv)

• Sampling 3D Gaussian Scenes in Seconds with Latent Diffusion Models (arXiv)

• Neural Appearance Modeling From Single Images (arXiv)

• GFFE: G-buffer Free Frame Extrapolation for Low-latency Real-time Rendering (arXiv)

Misc

• Luma has released a new feature that connects start and end keyframes for more AI video control. Look at these 10 wild examples (X)

• Nice. Maybe now we can stop pretending everyone isn’t using AI and get on with the business of making amazing films and allowing us AI filmmakers to show you what we can really do without pussyfooting around or ridiculing it (X)