Rethinking Diffusion Model-Based Video Super-Resolution: Leveraging Dense Guidance from Aligned Features
This work addresses video super-resolution for applications requiring high perceptual quality and fidelity, representing an incremental improvement over existing methods.
The paper tackled error accumulation and quality-fidelity trade-offs in diffusion model-based video super-resolution by proposing a method that uses aligned features for dense guidance, achieving a 35.82% reduction in DISTS, a 0.20 dB PSNR gain, and a 30.37% reduction in tLPIPS.
Diffusion model (DM) based Video Super-Resolution (VSR) approaches achieve impressive perceptual quality. However, they suffer from error accumulation, spatial artifacts, and a trade-off between perceptual quality and fidelity, primarily caused by inaccurate alignment and insufficient compensation between video frames. In this paper, within the DM-based VSR pipeline, we revisit the role of alignment and compensation between adjacent video frames and reveal two crucial observations: (a) the feature domain is better suited than the pixel domain for information compensation due to its stronger spatial and temporal correlations, and (b) warping at an upscaled resolution better preserves high-frequency information, but this benefit is not necessarily monotonic. Therefore, we propose a novel Densely Guided diffusion model with Aligned Features for Video Super-Resolution (DGAF-VSR), with an Optical Guided Warping Module (OGWM) to maintain high-frequency details in the aligned features and a Feature-wise Temporal Condition Module (FTCM) to deliver dense guidance in the feature domain. Extensive experiments on synthetic and real-world datasets demonstrate that DGAF-VSR surpasses state-of-the-art methods in key aspects of VSR, including perceptual quality (35.82\% DISTS reduction), fidelity (0.20 dB PSNR gain), and temporal consistency (30.37\% tLPIPS reduction).