Problem
Research questions and friction points this paper is trying to address.
Video Super-Resolution
Ill-posed Inverse Problem
Temporal Consistency
Long Video Processing
LoCoVSR: Local Context Diffusion Posterior Sampling for Video Super-Resolution
🤖 AI Summary
This study addresses the challenges of optical flow estimation errors and cumulative error propagation in recurrent processing for video super-resolution. To overcome these limitations, we propose an efficient reconstruction method based on a pixel-space denoising diffusion probabilistic model. By integrating diffusion posterior sampling with spatiotemporal local context learning and employing a moving average window mechanism to prevent information loss in latent space, our approach enables parallel inference for long videos without relying on optical flow estimation. Experimental evaluations on the VFHQ dataset demonstrate that the proposed method achieves high-fidelity and temporally consistent video super-resolution, significantly outperforming existing diffusion-based approaches.
📝 Abstract
Video super-resolution (VSR) is an ill-posed inverse problem that aims to reconstruct a high-resolution (HR) video from a noisy, low-resolution (LR) version of it. We present LoCoVSR, a diffusion-based VSR framework that leverages pixel-space denoising diffusion probabilistic models. LoCoVSR integrates the Diffusion Posterior Sampling technique with spatio-temporal context learning, operating in a moving-average form. A localized window of adjacent LR frames is used for recovering each center frame, while applying a shared noise trajectory across all frames. The localized windowing enables processing of long videos without length limitations, supports parallel inference, and prevents error accumulation that may occur in recursive processing. Unlike prior methods, LoCoVSR offers a simple yet very effective VSR solution, avoiding explicit optical flow estimation, or information loss caused by latent space processing. Trained on the VFHQ face dataset, LoCoVSR achieves accurate, temporally consistent and high-quality upscaling with competitive results against recent diffusion-based VSR approaches.
Innovation
Methods, ideas, or system contributions that make the work stand out.
Video Super-Resolution
Diffusion Posterior Sampling
Pixel-space Denoising
Spatio-temporal Context Learning
Local Windowing
🔎 Similar Papers
No similar papers found.
