New diffusion model techniques accelerate video restoration and image sampling

报道来源 [6]

1. arXiv cs.AI TIER_1 English(EN) · Taesung Kwon, Jonghyun Park, Hyungjin Chung, Jong Chul Ye · 2026-05-22 04:00

   Accelerating Video Inverse Problem Solvers with Autoregressive Diffusion Models

   arXiv:2605.20624v1 Announce Type: cross Abstract: Diffusion models provide powerful priors for zero-shot video inverse problems, but their real-time deployment is hindered by two inefficiencies: high initial latency caused by holistic video restoration, and low throughput resulti…

2. arXiv cs.LG TIER_1 English(EN) · Yichao Zhang · 2026-05-20 16:41

   Disentangling Generation and Regression in Stochastic Interpolants for Controllable Image Restoration

   Recent advances in Image Restoration (IR) have been largely driven by generative methods such as Diffusion Models and Flow Matching, which excel in synthesizing realistic textures while suffering from slow multi-step inference and compromised pixel fidelity. In contrast, classica…

3. arXiv cs.AI TIER_1 English(EN) · Jong Chul Ye · 2026-05-20 02:16

   Accelerating Video Inverse Problem Solvers with Autoregressive Diffusion Models

   Diffusion models provide powerful priors for zero-shot video inverse problems, but their real-time deployment is hindered by two inefficiencies: high initial latency caused by holistic video restoration, and low throughput resulting from multiple VAE passes to enforce measurement…

4. Hugging Face Daily Papers TIER_1 English(EN) · 2026-05-12 11:38

   Improving Diffusion Posterior Samplers with Lagged Temporal Corrections for Image Restoration

   Diffusion-based posterior sampling (PS) is a leading framework for imaging inverse problems, combining learned priors with measurement constraints. Yet, its standard formulations rely on instantaneous data-consistent estimates, which induce temporal variability in the reverse dyn…

5. arXiv stat.ML TIER_1 English(EN) · Van Khoa Nguyen, Lionel Blond\'e, Alexandros Kalousis · 2026-05-19 04:00

   Stein Diffusion Guidance: Training-Free Posterior Correction for Sampling Beyond High-Density Regions

   arXiv:2507.05482v3 Announce Type: replace-cross Abstract: Training-free diffusion guidance offers a flexible framework for leveraging off-the-shelf classifiers without additional training. Yet, current approaches hinge on posterior approximations via Tweedie's formula, which ofte…

6. arXiv cs.CV TIER_1 English(EN) · Nan Yang, Julian Straub, Fan Zhang, Richard Newcombe, Jakob Engel, Lingni Ma · 2026-05-08 04:00

   LAMP: Localization Aware Multi-camera People Tracking in Metric 3D World

   arXiv:2605.05390v1 Announce Type: new Abstract: Tracking 3D human motion from egocentric multi-camera headset is challenged by severe egomotion, partial visibility or occlusions and lack of training data. Existing methods designed for monocular video often require static or slowl…