基于扩散的视频超分辨率中，视频质量模型的准确度如何？

摘要

近期视频超分辨率（VSR）方法采用深度神经网络增强低质量输入视频并恢复视觉细节，其中基于扩散的方法展现出显著潜力。本文通过比较模型预测与主观测试结果，探究现有视频质量模型能否评估这些基于扩散的VSR方法性能。研究针对经压缩（AV1和DCVC-RT）与未压缩的低分辨率视频，在UHD-1/4K屏幕播放场景下，对比了六种超分辨率方法（Lanczos、Rhea、SCST、DOVE、SeedVR2、Starlight Mini）。采用一系列全参考与无参考质量模型评估其对这类新型质量退化的适用性，重点关注序列内性能。结果表明，基于CNN的全参考模型（如LPIPS、DISTS和CVQA-FR）的相关性系数显著高于传统全参考模型及所有测试的无参考模型。多数模型高估了SCST过度锐化的结果，而VMAF主要因Starlight Mini引入的空间不一致性导致性能失效。所有测试的视频质量模型均未达到足以替代补充性主观测试的精度。本文附带的原始、退化及超分辨率视频，以及用户评分与模型分数，均以开放数据形式发布至https://github.com/Telecommunication-Telemedia-Assessment/AVT-VQDB-UHD-1-VSR。

English

Recent video super-resolution (VSR) approaches use deep neural networks to enhance low-quality input videos and recover visual detail, with diffusion-based methods in particular showing promising results. In this paper, we investigate whether existing video quality models can be used to assess the performance of these diffusion-based VSR methods, by comparing model predictions with results from a subjective test. The study compares six upscaling methods (Lanczos, Rhea, SCST, DOVE, SeedVR2, Starlight Mini) applied to both compressed (AV1 and DCVC-RT) and uncompressed low-resolution videos considering the play-out on a UHD-1/4K screen. A range of full- and no-reference quality models are used to assess their applicability to this new type of quality degradation, focusing on within-sequence performance. The results highlight that CNN-based full-reference models, such as LPIPS, DISTS, and CVQA-FR show significantly higher correlation coefficients than both conventional full- as well as the tested no-reference models. Most overestimate the overly sharp results of SCST, with VMAF mainly failing due to spatial inconsistencies introduced by Starlight Mini. None of the tested video quality models reach sufficient accuracy so as to replace complementary subjective testing. The reference, degraded and upscaled videos, as well as the user ratings and model scores are made available with the paper at https://github.com/Telecommunication-Telemedia-Assessment/AVT-VQDB-UHD-1-VSR as open data.