IPI-VUGC: Acceptance/Annoyance and Video Quality of Vertically Oriented User Generated Videos

IPI-VUGC is a publicly available dataset of vertically oriented user generated video content that are typically found in online social media platforms. It consists of 48 unique SRCs compressed at 6 different rates, resulting in 336 videos in total. Two sets of experiments were conducted to collect both Quality of Experience labels in Acceptance/Annoyance paradigm and traditional video quality scores. The experiments were conducted in laboratory conditions at the facilities of Nantes University, France. All collected individual and mean opinion scores are shared with the community. More information regarding the dataset can be found in the ReadMe file. For further questions, please send an email to the author emails listed in the ReadMe File.

IPI-VUGC是一款公开可用的竖画幅用户生成视频数据集，此类内容广泛见于各类在线社交媒体平台。该数据集包含48个独立的SRC，经6种不同压缩码率处理后，最终共计得到336段视频。研究团队开展两组实验，分别采集接受度-烦扰度范式下的体验质量（Quality of Experience）标签，以及传统视频质量主观评分。实验均在法国南特大学的实验室设施中完成。所有采集到的个体主观评分与平均意见得分（Mean Opinion Score，MOS）均已向学术共同体公开共享。有关该数据集的更多详细信息，请参阅自述文件（ReadMe）。若有进一步疑问，请致信自述文件中列出的作者邮箱。