five

Nexdata/65_People_15204_Videos_of_Sports_and_Fitness_Video_Data

收藏
Hugging Face2024-04-11 更新2024-06-12 收录
下载链接:
https://hf-mirror.com/datasets/Nexdata/65_People_15204_Videos_of_Sports_and_Fitness_Video_Data
下载链接
链接失效反馈
官方服务:
资源简介:
--- license: cc-by-nc-nd-4.0 --- ## Description 65 People –15,204 Videos of Sports and Fitness Video Data. The data collection scene is indoor scenes. The race distribution is Asian, black and Caucasian; the age distribution is young and middle-aged people. The collection device is IR and RGB cameras. The dataset diversity includes different races, different age groups, different shooting angles, different collection distances, different human body orientations, different costumes and various fitness actions. The data can be used for tasks such as human behavior recognition and human segmentation in fitness scenes. For more details, please refer to the link: https://www.nexdata.ai/dataset/1212?source=Huggingface ## Data size 65 people, 15,204 videos ## Population distribution race distribution: 37 Asians, 24 Caucasians, 4 Black People; gender distribution: 24 males, 41 females; age distribution: 61 young people, 4 middle-aged people ## Collecting environment: indoor scenes ## Data diversity different races, different age groups, different shooting angles, different collection distances, different human body orientations, different costumes and various fitness actions ## Device infrared and color cameras, the camera resolution is 1,920x1,080 ## Collecting angle eye-level angle, simultaneous collection by three cameras (left, middle and right) ## Data format .mp4 ## Collection content collecting fitness videos of different people under one or two sets of clothing ## Accuracy the accuracy of video action is not less than 97%; the accuracy of label annotation is not less than 97% # Licensing Information Commercial License
提供机构:
Nexdata
原始信息汇总

数据集描述

65人,共15,204个体育和健身视频数据。数据采集场景为室内场景。种族分布包括亚洲人、黑人和高加索人;年龄分布为年轻人和中年人。采集设备为红外和RGB相机。数据集多样性包括不同种族、不同年龄组、不同拍摄角度、不同采集距离、不同人体方向、不同服装和各种健身动作。数据可用于健身场景中的人类行为识别和人体分割等任务。

数据规模

65人,15,204个视频

人口分布

  • 种族分布:37名亚洲人,24名高加索人,4名黑人
  • 性别分布:24名男性,41名女性
  • 年龄分布:61名年轻人,4名中年人

采集环境

室内场景

数据多样性

不同种族、不同年龄组、不同拍摄角度、不同采集距离、不同人体方向、不同服装和各种健身动作

设备

红外和彩色相机,相机分辨率为1,920x1,080

采集角度

平视角度,由三台相机(左、中、右)同时采集

数据格式

.mp4

采集内容

采集不同人在一套或两套服装下的健身视频

准确性

视频动作准确率不低于97%;标签标注准确率不低于97%

许可信息

商业许可

5,000+
优质数据集
54 个
任务类型
进入经典数据集
二维码
社区交流群

面向社区/商业的数据集话题

二维码
科研交流群

面向高校/科研机构的开源数据集话题

数据驱动未来

携手共赢发展

商业合作