OVIS_RGBD
收藏OVIS_RGBD 数据集概述
基本信息
- 数据集名称:OVIS_RGBD
- 许可证:apache-2.0
- 主要标签:segmentation
- 主要语言:en
数据集来源与目的
- 该数据集为论文《Beyond Appearance: Geometric Cues for Robust Video Instance Segmentation》而创建,论文链接为:https://arxiv.org/abs/2507.05948。
- 具体使用方法可在GitHub仓库(https://github.com/QuanzhuNiu/DVIS_Depth)中找到。
数据构成与处理
- 原始数据来源:原始帧图像和标注来自OVIS数据集(https://songbai.site/ovis/)。
- 深度信息生成:使用DepthAnythingV2(https://depth-anything-v2.github.io/)对所有图像进行单目深度估计。
- 数据格式:将深度图在通道维度上进行拼接,最终每张图像为RGBD格式。
相关引用
BibTeX @InProceedings{niu2025, author = {Niu, Quanzhu and Zhou, Yikang and Chen, Shihao and Zhang, Tao and Ji, Shunping}, title = {Beyond Appearance: Geometric Cues for Robust Video Instance Segmentation}, booktitle = {ICCV Workshops}, year = {2025}, } @InProceedings{OVIS, author = {Jiyang Qi and Yan Gao and Yao Hu and Xinggang Wang and Xiaoyu Liu and Xiang Bai and Serge Belongie and Alan Yuille and Philip Torr and Song Bai}, title = {Occluded Video Instance Segmentation: A Benchmark}, booktitle = {ICCV}, year = {2022}, } @InProceedings{depth_anything_v2, author = {Yang, Lihe and Kang, Bingyi and Huang, Zilong and Zhao, Zhen and Xu, Xiaogang and Feng, Jiashi and Zhao, Hengshuang}, title = {Depth Anything V2}, booktitle = {NeurIPS}, year = {2024}, }




