Long-RVOS：面向长时指代视频目标分割的综合基准
Long-RVOS: A Comprehensive Benchmark for Long-term Referring Video Object Segmentation
Long-RVOS 提供了长时指代视频目标分割的基准和评估指标，可直接用于测试模型在遮挡、消失重现和镜头切换下的时序建模能力；其基线 ReferMo 融合运动信息与局部到全局架构，为长视频 RVOS 提供了可复用的设计思路。
原文依据 we introduce Long-RVOS, a large-scale benchmark for long-term referring video object segmentation
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论指代视频目标分割（RVOS）旨在根据语言描述识别、跟踪并分割视频中的目标，近年来受到了广泛关注。然而，现有数据集仍集中于数秒内的短视频片段，且显著目标在大多数帧中可见。为将该任务推向更实际的应用场景，我们提出了 Long-RVOS，一个面向长时指代视频目标分割的大规模基准。Long-RVOS 包含 2000 多个平均时长超过 60 秒的视频，涵盖经历遮挡、消失-重现和镜头切换的多种目标。目标被人工标注了三种不同类型的描述，以分别评估对静态属性、运动模式以及时空关系的理解。此外，我们引入了两个新指标来评估时序一致性和时空一致性。我们在 Long-RVOS 上对 7 种最优方法进行了基准测试，结果表明当前方法在长视频挑战下表现严重不足。我们进一步提出了 ReferMo，一种有前景的基线方法，它融合运动信息以扩展时序感受野，并采用局部到全局的架构来同时捕捉短期动态和长期依赖。我们希望 Long-RVOS 和我们的基线能够推动未来的 RVOS 研究走向更真实和长格式的视频。我们的数据集和代码可在 https://isee-laboratory.github.io/Long-RVOS 获取。
英文原文摘要 1504 字符
Referring video object segmentation (RVOS) aims to identify, track and segment the objects in a video based on language descriptions, which has received great attention in recent years. However, existing datasets remain focus on short video clips within several seconds, with salient objects visible in most frames. To advance the task towards more practical scenarios, we introduce Long-RVOS, a large-scale benchmark for long-term referring video object segmentation. Long-RVOS contains 2,000+ videos of an average duration exceeding 60 seconds, covering a variety of objects that undergo occlusion, disappearance-reappearance and shot changing. The objects are manually annotated with three different types of descriptions to individually evaluate the understanding of static attributes, motion patterns and spatiotemporal relationships. Moreover, we introduce two new metrics to assess the temporal and spatiotemporal consistency. We benchmark 7 state-of-the-art methods on Long-RVOS to show that current approaches struggle severely with the long-video challenges. We further propose ReferMo, a promising baseline method that integrates motion information to expand the temporal receptive field, and employs a local-to-global architecture to capture both short-term dynamics and long-term dependencies. We hope that Long-RVOS and our baseline can drive future RVOS research towards more realistic and long-form videos. Our dataset and code is available at https://isee-laboratory.github.io/Long-RVOS.