Exploring Video Quality Assessment on User Generated Contents from Aesthetic and Technical Perspectives

By Haoning Wu, Erli Zhang, Liang Liao, Chaofeng Chen, Jingwen Hou, Annan Wang, Wenxiu Sun, Qiong Yan, Weisi Lin. IEEE International Conference on Computer Vision, 2022

427 citations recorded by the source corpus.

The rapid increase in user-generated content (UGC) videos calls for the development of effective video quality assessment (VQA) algorithms. However, the objective of the UGC-VQA problem is still ambiguous and can be viewed from two perspectives: the $\color{Green}{\text{technical perspective}}$, measuring the perception of distortions; and the $\color{Blue}{\text{aesthetic perspective}}$, which relates to preference and recommendation on contents. To understand how these two perspectives affect overall subjective opinions in UGC-VQA, we conduct a large-scale subjective study to collect human quality opinions on the overall quality of videos as well as perceptions from aesthetic and technical perspectives. The collected Disentangled Video Quality Database (DIVIDE-3k) confirms that human quality opinions on UGC videos are universally and inevitably affected by both aesthetic and technical perspectives. In light of this, we propose the Disentangled Objective Video Quality Evaluator (DOVER) to learn the quality of UGC videos based on the two perspectives. The DOVER proves state-of-the-art performance in UGC-VQA under very high efficiency. With perspective opinions in DIVIDE-3k, we further propose DOVER++, the first approach to provide reliable clear-cut quality evaluations from a single aesthetic or technical perspective. Code at https://github.com/VQAssessment/DOVER.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/b8ac29f2da80cf3f7f367cf6ebabe3b18114a2e6

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

1. Ying Shan
2. Xintao Wang
3. Xiaodong Cun
4. Yong Zhang
5. Haoning Wu
6. Liang Liao
7. Erli Zhang
8. Annan Wang

Citations in the corpus

• HunyuanVideo: A Systematic Framework For Large Video Generative Models (2024, positive): “We employ Dover [85] to assess the visual aesthetics of video clips from both aesthetic and technical viewpoints.”
• Q-Align: Teaching LMMs for Visual Scoring via Discrete Text-Defined Levels (2023, background): “Named as video quality assessment (VQA), the focus of this task is also kind of complicated, that several studies have claimed that scores are not only affected by quality issues, but also contents (Li et al., 2019), and even aesthetics (Wu et al., 2023d).”
• EvalCrafter: Benchmarking and Evaluating Large Video Generation Models (2023, methodology): “We utilize the state-of-the-art video quality assessment method, Dover [60], to evaluate the quality of the generated video in terms of aes-thetics and technicality, where the technical rating measures the quality of the generated video in terms of the common distortion, including noises,…”
• VideoCrafter2: Overcoming Data Limitations for High-Quality Video Diffusion Models (2024, methodology): “The technical and aesthetic scores in Eval-Crafter are adapted from DOVER [54].”
• OpenVid-1M: A Large-Scale High-Quality Dataset for Text-to-video Generation (2024, neutral): “Additionally, OpenVid - 1M integrates the recent model DOVER [29] to assess the video clarity, preserving high-quality videos with clean textures.”