Tag: denoising diffusion models
Browse our exclusive articles!
FVD: Efficient Inference Alignment for Diffusion Models
Discover FVD, a novel Fleming-Viot resampling method improving diffusion model alignment, boosting diversity, speed, and image generation quality.
Selective Attention Aggregation Boosts Diffusion Visuals
Enhance diffusion-based visual interpretation with selective aggregation of attention maps, improving accuracy and model transparency in text-to-image gene...
Robust Diffusion-Based Image Compression vs Bit-Flip Errors
Explore how diffusion-based image compression resists bit-flip errors, ensuring high-quality visuals even over unreliable channels using RCC frameworks.
Rectified Schrödinger Bridge for Fast Visual Navigation
Discover Rectified Schrödinger Bridge Matching for efficient few-step visual navigation, boosting autonomous agent performance with minimal integration ste...
FastDiSS: Efficient Few-Step Diffusion for Seq2Seq Models
Discover FastDiSS, a fast and robust diffusion language model enhancing sequence-to-sequence generation with fewer steps and higher accuracy.