This weekend in Generative Media
Google's AI search furor; Other celebs face AI voice clones; Golden Gate Claude
Google’s A.I. Search Errors Cause a Furor Online (New York Times)
Here’s Why Other Celebrities Could Face Problems With AI Voice Cloning—Not Just Scarlett Johansson (Forbes)
This week, we showed how altering internal "features" in our AI, Claude, could change its behavior. We found a feature that can make Claude focus intensely on the Golden Gate Bridge. Now, for a limited time, you can chat with Golden Gate Claude: (X) http://claude.ai
StopThePop: Sorted Gaussian Splatting for View-Consistent Real-time Rendering (SIGGRAPH 2024, project page)
Flexible Motion In-betweening with Diffusion Models (SIGGRAPH 2024, project page)
ReVideo: Remake a Video with Motion and Content Control (project page)
Generative Camera Dolly: Extreme Monocular Dynamic Novel View Synthesis (project page)
NeRF-Casting: Improved View-Dependent Appearance with Consistent Reflections (project page)
Neural Directional Encoding for Efficient and Accurate View-Dependent Appearance Modeling (project page)
InstaDrag: Lightning Fast and Accurate Drag-based Image Editing Emerging from Videos (project page)
SignLLM: Sign Languages Production Large Language Models (project page)
EditWorld: Simulating World Dynamics for Instruction-Following Image Editing (arXiv) Code on GitHub.
Tele-Aloha: A Low-budget and High-authenticity Telepresence System Using Sparse RGB Cameras (SIGGRAPH 2024, arXiv)
"Previously on ..." From Recaps to Story Summarization (CVPR 2024, arXiv)
TexSliders: Diffusion-Based Texture Editing in CLIP Space (SIGGRAPH 2024, arXiv)
RectifID: Personalizing Rectified Flow with Anchored Classifier Guidance (arXiv) Code on GitHub.
SketchDream: Sketch-based Text-to-3D Generation and Editing (arXiv)
CamViG: Camera Aware Image-to-Video Generation with Multimodal Transformers (arXiv)