FudanCVL/MOSEv2

MOSEv2是一个全面的电影对象分割数据集，旨在在现实世界条件下推动VOS方法的发展。它包含5024部电影，超过701,976个高质量掩码，涉及10,074个对象，跨越200个类别。与现有数据集相比，MOSEv2引入了更大的场景复杂性，包括更频繁的对象消失和重现、严重遮挡和拥挤、小对象、恶劣天气条件、低光场景、多镜头序列、伪装对象、非物理目标以及需要外部知识 scenarios。

MOSEv2 is a comprehensive video object segmentation dataset designed to advance VOS methods under real-world conditions. It consists of 5,024 videos and over 701,976 high-quality masks for 10,074 objects across 200 categories. MOSEv2 introduces significantly greater scene complexity compared to existing datasets, including more frequent object disappearance and reappearance, severe occlusions and crowding, smaller objects, adverse weather conditions, low-light scenes, multi-shot sequences, camouflaged objects, non-physical targets, and scenarios requiring external knowledge.