UNISV-Dataset
收藏资源简介:
介绍了一个包含1,200个样本的夜间红外监控视频行为识别数据集。该数据集使用录制的、未经编辑的原始视频数据作为样本,涵盖10种不同的行为动作。每个动作类别包含120个样本。数据集中的视频样本命名遵循UCF-101数据集的格式,例如,视频样本名可能是v_DoubleWave_g01_c01,其中v_DoubleWave表示双挥手的行为类别,g01表示该视频样本是在位置01录制的,c01表示在位置01的第一个双挥手行为类别的视频样本。所有视频样本均在户外环境中录制,场景选择基于现实监控摄像头的关键位置,如停车场、花园、小巷等。所有视频样本的录制时间均为夜间,数据集编译涉及15名身高和体型各异的实验参与者。
A dataset for nighttime infrared surveillance video behavior recognition, comprising 1,200 samples, is introduced. The dataset utilizes raw, unedited recorded video data as samples, encompassing 10 distinct behavioral actions. Each action category includes 120 samples. The video samples in the dataset are named following the UCF-101 dataset format; for example, a video sample might be named v_DoubleWave_g01_c01, where 'v_DoubleWave' denotes the behavior category of double waving, 'g01' indicates that the video sample was recorded at location 01, and 'c01' signifies the first video sample of the double waving behavior category at location 01. All video samples were recorded in outdoor environments, with scene selections based on key locations typical of real surveillance cameras, such as parking lots, gardens, and alleys. All video samples were recorded at night, and the dataset compilation involved 15 experimental participants of varying heights and body types.
UNISV-Dataset 概述
数据集描述
- 样本数量:包含1,200个样本。
- 行为类别:涵盖10种不同的行为动作,每个类别包含120个样本。
- 视频特性:所有视频均为未编辑或随机编辑的原始视频,大部分视频不包含人物。视频命名遵循UCF-101格式,例如"v_DoubleWave_g01_c01",其中"v_DoubleWave"表示行为类别,"g01"和"c01"分别代表地点和样本编号。
- 场景设置:所有视频均在户外夜间录制,场景选择基于实际监控摄像头的关键位置,如停车场、花园、小巷等。
- 参与者:数据集涉及15名实验参与者,具有不同的身高和体型。
行为类别
- 类别列表:"Double Wave," "Wave Hand," "Walk," "Jump," "Squat," "Jogging," "Push People," "Shake Hands," "Embrace," 和 "Fight"。
- 参与者数量:每个行为类别涉及6名参与者,总共15名参与者。
- 视频样本数:每个行为类别有120个视频样本,总计1200个视频样本。
- 视频规格:视频帧率为10帧每秒,分辨率为480 x 248像素。
引用信息
若使用此数据集进行研究,请引用:
Feng Z, Wang X, Zhou J, et al. MDJ: A Multi-Scale Difference Joint Keyframe Extraction Algorithm for Infrared Surveillance Video Action Recognition[J]. Digital Signal Processing, 2024: 104469. https://doi.org/10.1016/j.dsp.2024.104469.




