YouTube UGC Dataset for Video Compression Research
Open accessAbstract
Non-professional video, commonly known as User Generated Content (UGC) has become very popular in today's video sharing applications. However, traditional metrics used in compression and quality assessment, like BD-Rate and PSNR, are designed for pristine originals. Thus, their accuracy drops significantly when being applied on non-pristine originals (the majority of UGC). Understanding difficulties for compression and quality assessment in the scenario of UGC is important, but there are few public UGC datasets available for research. This paper introduces a large scale UGC dataset (1500 20 sec video clips) sampled from millions of YouTube videos. The dataset covers popular categories like Gaming, Sports, and new features like High Dynamic Range (HDR). Besides a novel sampling method based on features extracted from encoding, challenges for UGC compression and quality evaluation are also discussed. Shortcomings of traditional reference-based metrics on UGC are addressed. We demonstrate a promising way to evaluate UGC quality by no-reference objective quality metrics, and evaluate the current dataset with three no-reference metrics (Noise, Banding, and SLEEQ).
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
YouTube-UGC Dataset comprises 1,380 20-second video clips sampled from millions of YouTube videos, rated by more than 8,000 human subjects.
“The most recently published UGC-VQA database is the YouTube-UGC Dataset [11] comprising 1,380 20-second video clips sampled from millions of YouTube videos, which were rated by more than 8,000 human subjects.”
From UGC-VQA: Benchmarking Blind Video Quality Assessment for User Generated Content (2021) · cites this paper as [11]
Cites
Image quality assessment: from error visibility to structural similarity
Wang, Z., Bovik, A. C., Sheikh, H. R., & Simoncelli, E. P. (2004) · IEEE Transactions on Image Processing · doi:10.1109/tip.2003.819861 · Cited by 57,892 · see citing works
Large-scale cluster management at Google with Borg
Verma, A., Pedrosa, L., Korupolu, M., Oppenheimer, D., Tune, E., & Wilkes, J. (2015) · Proceedings of the Tenth European Conference on Computer Systems · doi:10.1145/2741948.2741964 · Cited by 1,468 · cited 1× · see citing works
“LIVE datasets [10]-[ 12] are also quite popular.”
Flagged first, then by citations
Cited by
Sorted by citations
UGC-VQA: Benchmarking Blind Video Quality Assessment for User Generated Content
Tu, Z., Wang, Y., Birkbeck, N., Adsumilli, B., & Bovik, A. C. (2021) · IEEE Transactions on Image Processing · doi:10.1109/tip.2021.3072221 · Cited by 363 · see citing works
“PUBLIC LARGE-SCALE USER-GENERATED CONTENT VIDEO QUALITY ASSESSMENT (UGC-VQA) DATABASES COMPARED: KONVID-1K [10], LIVE-VQC [9], AND YOUTUBE-UGC [11]”
BackgroundMethodInfluentialSelf-citationTu, Z., Lin, J., Wang, Y., Adsumilli, B., & Bovik, A. C. (2020) · IEEE Signal Processing Letters · doi:10.1109/lsp.2020.3024985 · Cited by 31 · see citing works
Self-citationWhy no-reference metrics for image and video quality lack accuracy and reproducibility
Pinson, M. H. (2023) · IEEE Transactions on Broadcasting · doi:10.1109/tbc.2022.3191059 · Cited by 19 · see citing works
- Authors
- Wang, Y., Inguva, S., & Adsumilli, B.
- Affiliations
- Google (United States) (US)
- Venue
- 2019 IEEE 21st International Workshop on Multimedia Signal Processing (MMSP)
- Publisher
- IEEE
- Published
- 2019-09
- Pages
- 1-5
- OpenAlex
- W2939995367
- Cited by
- 282
- References
- 2
- Citations by year
- 2026: 25 · 2025: 55 · 2024: 55 · 2023: 50 · 2022: 43 · 2021: 42 · 2020: 12
- FWCI
- 7.91
- Type
- conference paper
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- IEEE International Workshop on Multimedia Signal Processing (MMSP) (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Image and Video Quality Assessment
- Physical Sciences › Computer Science › Signal Processing › Video Coding and Compression Technologies
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Data Compression Techniques
- Keywords
- user-generated content, video compression, YouTube, no-reference quality metrics, high dynamic range video, video quality assessment
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
