---
title: SegTrack-v2
canonical_url: "https://www.modelscope.cn/datasets/OpenDataLab/SegTrack-v2"
md_url: "https://www.modelscope.cn/datasets/OpenDataLab/SegTrack-v2.md"
repository: OpenDataLab/SegTrack-v2
last_updated: 2025-02-27
license: "[Unknown]"
storage_size: "200 MB"
domain:
  - publishDate
  - publisher
  - publishUrl
  - paperUrl
  - mediaTypes
  - displayName
  - labelTypes
  - taskTypes
tasks:
  - 2013
  - "Georgia Institute of Technology"
  - "https://web.engr.oregonstate.edu/~lif/SegTrack2/dataset.html"
  - "https://web.engr.oregonstate.edu/~lif/SegTrack2/segtrack2_cameraready.pdf"
  - Image
  - SegTrack-v2
  - SemanticSegMap
  - "Video Semantic Segmentation"
downloads: 3296
stars: 0
---

# SegTrack-v2

> SegTrack-v2 - OpenDataLab 在 ModelScope 开源的数据集。displayName: SegTrack-v2 labelTypes: SemanticSegMap license: Unknown mediaTypes: Video Image paperUrl: https://web.engr.oregonstate.edu/lif/SegTrack2/segtrack2cameraready.pdf publishDate: "2013" publishUrl:…

OpenDataLab/SegTrack-v2 是 ModelScope 魔搭社区上的2013、Georgia Institute of Technology、https://web.engr.oregonstate.edu/~lif/SegTrack2/dataset.html数据集，涉及 publishDate、publisher、publishUrl 领域，存储大小 200 MB，采用 [Unknown] 许可。

- **Repository**: OpenDataLab/SegTrack-v2
- **License**: [Unknown]
- **Tasks**: 2013, Georgia Institute of Technology, https://web.engr.oregonstate.edu/~lif/SegTrack2/dataset.html, https://web.engr.oregonstate.edu/~lif/SegTrack2/segtrack2_cameraready.pdf, Image, SegTrack-v2, SemanticSegMap, Video Semantic Segmentation
- **Domain**: publishDate, publisher, publishUrl, paperUrl, mediaTypes, displayName, labelTypes, taskTypes
- **Storage size**: 200 MB
- **Downloads**: 3296
- **Stars**: 0
- **Last updated**: 2025-02-27

Source: https://www.modelscope.cn/datasets/OpenDataLab/SegTrack-v2

---

displayName: SegTrack-v2
labelTypes:
- SemanticSegMap
license:
- Unknown
mediaTypes:
- Video
- Image
paperUrl: https://web.engr.oregonstate.edu/~lif/SegTrack2/segtrack2_cameraready.pdf
publishDate: "2013"
publishUrl: https://web.engr.oregonstate.edu/~lif/SegTrack2/dataset.html
publisher:
- Georgia Institute of Technology
tags:
- People
- Car
- Parachute
taskTypes:
- Video Object Segmentation
- Semantic Segmentation
- Unsupervised Video Object Segmentation
- Video Segmentation
- Video Semantic Segmentation

---
# 数据集介绍
  ## 简介
   SegTrack v2 是一个视频分割数据集，在每个视频的每一帧的多个对象上都有完整的像素级注释。 
  ## 引文
  
```
"@inproceedings{li2013video,
title={Video segmentation by tracking many figure-ground segments},
author={Li, Fuxin and Kim, Taeyoung and Humayun, Ahmad and Tsai, David and Rehg, James M},
booktitle={Proceedings of the IEEE international conference on computer vision},
pages={2192--2199},
year={2013}
}"
```

  
## Download dataset
:modelscope-code[]{type="git"}