NVIDIA paper: Posterior Matching Distillation for few-step diffusion image generation
nvidia · hf · 2026-11-20
A Virginia Tech × NVIDIA paper (arXiv, Nov 19) proposes Posterior Matching Distillation, which trains a stochastic student to match a teacher diffusion model's full posterior via tilted mean constraints, enabling high-quality few-step image generation without auxiliary score networks. Project page: pmd-distillation.github.io.
More from Multimodal
- Astra 6.2 achieves unprecedented 3D realism in image generation with extra steering — gabrielchua · 2026-11-20
- Nano Banana Pro 2 staged for release at Google, leak claims — mark_k · 2026-11-20
- 开源 MCP 服务器让 Claude 写 Strudel 代码现场作曲并驱动真实 MIDI 合成器 — konni_away · 2026-11-20
- 'Shadows': eerie AI short where a man's shadow moves on its own — ainsoph00 · 2026-11-20
- Stylized AI character designs shown side by side in 2D art and 3D renders — azed_ai · 2026-11-20
- Open-source RGBMocap adds live streaming to drive Unreal MetaHumans in real time — camenduru · 2026-11-20