VDPVE
收藏资源简介:
VDPVE数据集由上海交通大学图像通信与网络工程研究所构建,包含1211个经过不同增强处理的视频,旨在评估视频增强质量与人类视觉感知的匹配度。数据集分为三个子集:第一个子集包含600个视频,经过色彩、亮度和对比度增强;第二个子集包含310个视频,进行去模糊处理;第三个子集包含301个视频,进行去抖动处理。数据集通过邀请21名受试者进行主观评分,以获取每视频的平均意见分数(MOS)。VDPVE不仅推动了视频质量评估方法的发展,也在一定程度上提升了视频增强技术的性能。
The VDPVE dataset was constructed by the Institute of Image Communication and Network Engineering at Shanghai Jiao Tong University. It comprises 1,211 videos subjected to various enhancement treatments, aiming to evaluate the consistency between video enhancement quality and human visual perception. The dataset is divided into three subsets: the first subset includes 600 videos processed for color, brightness and contrast enhancements; the second subset contains 310 videos treated with deblurring; the third subset consists of 301 videos stabilized via de-jittering. Subjective scoring was conducted by inviting 21 participants to obtain the Mean Opinion Score (MOS) for each individual video. The VDPVE dataset not only promotes the development of video quality assessment methodologies, but also improves the performance of video enhancement technologies to a certain extent.




