UHD-GPGNet: UHD Video Denoising via Gaussian-Process-Guided Local Spatio-Temporal Modeling

TL;DR

UHD-GPGNet is a novel 4K video denoising framework that explicitly models local degradation and uncertainty using Gaussian processes, enabling real-time, high-quality denoising with fewer parameters.

Contribution

The paper introduces a Gaussian-process-guided local spatio-temporal model for UHD video denoising that explicitly characterizes degradation, enabling efficient and robust full-resolution inference.

Findings

Achieves competitive denoising fidelity with fewer parameters.

Enables real-time 4K inference with significant speedup.

Generalizes well to real sensor noise and improves object detection.

Abstract

Ultra-high-definition (UHD) video denoising requires simultaneously suppressing complex spatio-temporal degradations, preserving fine textures and chromatic stability, and maintaining efficient full-resolution 4K deployment. In this paper, we propose UHD-GPGNet, a Gaussian-process-guided local spatio-temporal denoising framework that addresses these requirements jointly. Rather than relying on implicit feature learning alone, the method estimates sparse GP posterior statistics over compact spatio-temporal descriptors to explicitly characterize local degradation response and uncertainty, which then guide adaptive temporal-detail fusion. A structure-color collaborative reconstruction head decouples luminance, chroma, and high-frequency correction, while a heteroscedastic objective and overlap-tiled inference further stabilize optimization and enable memory-bounded 4K deployment. Experiments on UVG and RealisVideo-4K show that UHD-GPGNet achieves competitive restoration fidelity with substantially fewer parameters than existing methods, enables real-time full-resolution 4K inference with significant speedup over the closest quality competitor, and maintains robust performance across a multi-level mixed-degradation schedule.A real-world study on phone-captured 4K video further confirms that the model, trained entirely on synthetic degradation, generalizes to unseen real sensor noise and improves downstream object detection under challenging conditions.