Back to datasets
Dataset assetOpen Source CommunityVideo Quality AssessmentVideo Classification
teowu/LSVQ-videos
This is an unofficial copy of the LSVQ dataset for no‑reference video quality assessment (NR‑VQA). Since the original dataset link is unavailable, this copy aims to facilitate related research. The dataset copyright belongs to Facebook Research and the LIVE Laboratory at the University of Texas at Austin.
Source
hugging_face
Created
Nov 28, 2025
Updated
Nov 13, 2023
Signals
348 views
Availability
Linked source ready
Overview
Dataset description and usage context
Dataset Overview
Basic Information
- License: MIT
- Task Category: Video Classification
- Tags: Video Quality Assessment
Dataset Description
- The dataset is an unofficial replica of the LSVQ dataset (Ying et al., CVPR 2021) for NR‑VQA.
- Because the original download link is no longer functional, this copy is provided to support research.
Label Files
- Training set labels:
train_labels.txt - Test subset labels:
labels_test.txt(LSVQ_test subset)labels_1080p.txt(LSVQ_1080p subset)
Copyright Information
- The dataset is copyrighted by Facebook Research and the LIVE Laboratory at UT Austin.
- If the copyright holders request removal, the unofficial repository may be taken down.
References
- Z. Ying, M. Mandal, D. Ghadiyaram, A. C. Bovik, “Patch‑VQ: ‘Patching Up’ the Video Quality Problem,” arXiv 2020.
- Z. Ying, M. Mandal, D. Ghadiyaram, A. C. Bovik, “LIVE Large‑Scale Social Video Quality (LSVQ) Database,” Online: https://github.com/baidut/PatchVQ, 2020.
Need downstream help?
Pair the dataset with AI analysis and content workflows.
Once the source passes your review, move straight into summarization, transformation, report drafting, or presentation generation with the JuheAI toolchain.