Déjà View is a new transformer model for multi-view 3D reconstruction that replaces stacked unique layers with a single transformer block applied recurrently at inference time. ...
Researchers from ETH Zurich, Google, Meta, and Athlence Sports have introduced PaGeR (Panoramic Geometry Reconstruction), a framework that adapts perspective-image 3D foundation...
Researchers from Disney Research|Studios and ETH Zurich have introduced RenderFlow, an end-to-end neural rendering framework that replaces conventional physically based renderin...
·
Researchers have proposed MUCS (Mirrored Unlearning and Noise-Consistent Skew), a new approach to training data attribution (TDA) for diffusion models, addressing longstanding r...
·
Stanford researchers have introduced Spectral Progressive Diffusion, a framework that exploits a known property of diffusion models — that low-frequency content is generated ear...
Script analysis has long been the province of experienced producers, story editors, and development executives who bring emotional intelligence, cultural context, and creative v...
·
Segmind has unveiled SegMoE (Segmind Mixture of Diffusion Experts), a new approach that applies the Mixture of Experts (MoE) paradigm to diffusion-based image generation models....
·
Stability AI has optimized Stable Diffusion XL inference using JAX on Google Cloud's TPU v5e hardware, targeting meaningfully faster image generation throughput. The work focuse...
·
InstructPix2Pix introduces a method for fine-tuning Stable Diffusion so that users can edit images using plain natural language commands — for example, "make the sky look stormy...
·
CLIPSeg brings together CLIP's powerful vision-language understanding with image segmentation, enabling pixel-level semantic analysis without requiring task-specific training da...
·
