SportsMOT

# Dataset Card for SportsMOT

<!-- Provide a quick summary of the dataset. -->

## Dataset Details

### Dataset Description

<!-- Provide a longer summary of what this dataset is. -->
Multi-object tracking (MOT) is a fundamental task in computer vision, aiming to estimate objects (e.g., pedestrians and vehicles) bounding boxes and identities in video sequences. We propose a large-scale multi-object tracking dataset named SportsMOT, consisting of 240 video clips from 3 categories (i.e., basketball, football and volleyball). The objective is to only track players on the playground (i.e., except for a number of spectators, referees and coaches) in various sports scenes.

### Dataset Sources [optional]

<!-- Provide the basic links for the dataset. -->

- **Repository:** https://github.com/MCG-NJU/SportsMOT
- **Paper:** https://arxiv.org/abs/2304.05170
- **Competiton:** https://codalab.lisn.upsaclay.fr/competitions/12424
- **Point of Contact:** mailto: [e-mail]

## Dataset Structure

<!-- This section provides a description of the dataset fields, and additional information about the dataset structure such as criteria used to create the splits, relationships between data points, etc. -->

Data in SportsMOT is organized in the form of MOT Challenge 17.

```
splits_txt(video-split mapping)
- basketball.txt
- volleyball.txt
- football.txt
- train.txt
- val.txt
- test.txt
scripts
- mot_to_coco.py
- sportsmot_to_trackeval.py
dataset(in MOT challenge format)
- train
- VIDEO_NAME1
- gt
- img1
- 000001.jpg
- 000002.jpg
- seqinfo.ini
- val(the same hierarchy as train)
- test
- VIDEO_NAME1
- img1
- 000001.jpg
- 000002.jpg
- seqinfo.ini
```

## Dataset Creation

### Curation Rationale

<!-- Motivation for the creation of this dataset. -->
Multi-object tracking (MOT) is a fundamental task in computer vision, aiming to estimate objects (e.g., pedestrians and vehicles) bounding boxes and identities in video sequences.

Prevailing human-tracking MOT datasets mainly focus on pedestrians in crowded street scenes (e.g., MOT17/20) or dancers in static scenes (DanceTrack). In spite of the increasing demands for sports analysis, there is a lack of multi-object tracking datasets for a variety of sports scenes, where the background is complicated, players possess rapid motion and the camera lens moves fast.

### Source Data

<!-- This section describes the source data (e.g. news text and headlines, social media posts, translated sentences, ...). -->
> We select three worldwide famous sports, football, basketball, and volleyball, and collect videos of high-quality professional games including NCAA, Premier League, and Olympics from MultiSports, which is a large dataset in sports area focusing on spatio-temporal action localization.

#### Annotation process

<!-- This section describes the annotation process such as annotation tools used in the process, the amount of data annotated, annotation guidelines provided to the annotators, interannotator statistics, annotation validation, etc. -->

We annotate the collected videos according to the following guidelines.

1. The entire athlete’s limbs and torso, excluding any other objects like balls touching the athlete’s body, are required to be annotated.

2. The annotators are asked to predict the bounding box of the athlete in the case of occlusion, as long as the athletes have a visible part of body. However, if half of the athletes’ torso is outside the view, annotators should just skip them.

3. We ask the annotators to confirm that each player has a unique ID throughout the whole clip.

### Dataset Curators

Authors of [SportsMOT: A Large Multi-Object Tracking Dataset in Multiple Sports Scenes](https://arxiv.org/pdf/2304.05170)

- Yutao Cui

- Chenkai Zeng

- Xiaoyu Zhao

- Yichun Yang

- Gangshan Wu

- Limin Wang

## Citation Information

<!-- If there is a paper or blog post introducing the dataset, the APA and Bibtex information for that should go in this section. -->

If you find this dataset useful, please cite as

```
@inproceedings{cui2023sportsmot,
title={Sportsmot: A large multi-object tracking dataset in multiple sports scenes},
author={Cui, Yutao and Zeng, Chenkai and Zhao, Xiaoyu and Yang, Yichun and Wu, Gangshan and Wang, Limin},
booktitle={Proceedings of the IEEE/CVF International Conference on Computer Vision},
pages={9921--9931},
year={2023}
}
```

# SportsMOT 数据集卡片

<!-- 提供数据集的简要概述。 -->

## 数据集详情

### 数据集描述

<!-- 详细说明该数据集的具体内容。 -->
多目标跟踪（Multi-object Tracking, MOT）是计算机视觉领域的基础任务，旨在估算视频序列中目标（如行人、车辆）的边界框与身份信息。本文提出了一个大规模多目标跟踪数据集SportsMOT，该数据集包含来自篮球、足球、排球3个类别的240段视频片段，其核心任务为在各类体育场景中仅跟踪赛场内的运动员（即排除部分观众、裁判与教练）。

### 数据集来源 [可选]

<!-- 提供数据集的基础链接。 -->

- **仓库地址**：https://github.com/MCG-NJU/SportsMOT
- **论文地址**：https://arxiv.org/abs/2304.05170
- **竞赛页面**：https://codalab.lisn.upsaclay.fr/competitions/12424
- **联系方式**：mailto: [e-mail]

## 数据集结构

<!-- 本章节介绍数据集的字段信息，以及划分依据、数据点间关系等额外结构相关内容。 -->

SportsMOT数据集采用MOT Challenge 17格式进行组织。

splits_txt(视频-划分映射文件)
- basketball.txt
- volleyball.txt
- football.txt
- train.txt
- val.txt
- test.txt
scripts
- mot_to_coco.py
- sportsmot_to_trackeval.py
dataset(采用MOT Challenge格式)
- train
- VIDEO_NAME1
- gt
- img1
- 000001.jpg
- 000002.jpg
- seqinfo.ini
- val(与train目录结构一致)
- test
- VIDEO_NAME1
- img1
- 000001.jpg
- 000002.jpg
- seqinfo.ini

## 数据集构建

### 编撰依据

<!-- 阐述创建该数据集的动机。 -->
多目标跟踪（MOT）是计算机视觉领域的基础任务，旨在估算视频序列中目标（如行人、车辆）的边界框与身份信息。

当前主流的人体跟踪MOT数据集主要聚焦于拥挤街景中的行人（如MOT17/20）或静态场景中的舞者（DanceTrack）。尽管体育分析的需求日益增长，但现有数据集仍缺乏针对各类体育场景的多目标跟踪数据集——这类场景往往背景复杂、运动员动作迅捷且相机快速移动。

### 源数据

<!-- 本章节描述源数据的相关信息（如新闻文本与标题、社交媒体帖文、译句等）。 -->

> 我们选取了足球、篮球、排球这三项全球知名的体育运动，从专注于时空动作定位的体育领域大型数据集MultiSports中采集了包括NCAA赛事、英超联赛以及奥运会在内的高质量职业比赛视频。

#### 标注流程

<!-- 本章节描述标注流程，如使用的标注工具、标注数据量、提供给标注人员的标注指南、标注者间一致性统计、标注验证等内容。 -->

我们按照以下指南对采集到的视频进行标注：

1. 需标注运动员完整的四肢与躯干，排除所有接触运动员身体的其他物体（如球类）。
2. 当运动员被遮挡时，只要其身体仍有可见部分，标注人员需估算该运动员的边界框；但若运动员躯干的一半超出画面范围，则应跳过该目标。
3. 要求标注人员确保每位运动员在整段视频片段中拥有唯一的身份标识（ID）。

### 数据集编撰者

论文《SportsMOT: A Large Multi-Object Tracking Dataset in Multiple Sports Scenes》（https://arxiv.org/pdf/2304.05170）作者
- Yutao Cui
- Chenkai Zeng
- Xiaoyu Zhao
- Yichun Yang
- Gangshan Wu
- Limin Wang

## 引用信息

<!-- 若存在介绍该数据集的论文或博客文章，需在此处提供APA与Bibtex格式的引用信息。 -->

若您使用本数据集，请引用以下文献：

@inproceedings{cui2023sportsmot,
title={Sportsmot: A large multi-object tracking dataset in multiple sports scenes},
author={Cui, Yutao and Zeng, Chenkai and Zhao, Xiaoyu and Yang, Yichun and Wu, Gangshan and Wang, Limin},
booktitle={Proceedings of the IEEE/CVF International Conference on Computer Vision},
pages={9921--9931},
year={2023}
}