How Accurate are Video Quality Models for Diffusion-Based Video Super-Resolution?
2026-05-25 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors studied how well current video quality models can judge the performance of new video enhancement methods that use diffusion techniques. They tested six different upscaling methods on both compressed and uncompressed videos shown on a 4K screen and compared model scores to human opinions. They found that some deep-learning based models, like LPIPS and DISTS, correlate better with human judgment than older or no-reference models, but none were accurate enough to fully replace human testing. They also shared all their data and results openly for others to use.
video super-resolutiondiffusion methodsvideo quality assessmentfull-reference modelsno-reference modelsLPIPSDISTSVMAFupscalingsubjective testing
Authors
Benjamin Herb, Steve Göring, Alexander Raake, Rakesh Rao Ramachandra Rao
Abstract
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.