NJUPT Gait dataset

Citation Author(s):
Shu
Shen
Nanjing University of Posts and Telecommunications
Wen-Juan
Li
Nanjing University of Posts and Telecommunications
Wei-Nan
Chen
Nanjing University of Posts and Telecommunications
Shao-Shan
Sun
Nanjing University of Posts and Telecommunications
Submitted by:
Shu Shen
Last updated:
Wed, 07/14/2021 - 23:36
DOI:
10.21227/3n2q-w610
Data Format:
License:
23 Views
Categories:
Keywords:
0
0 ratings - Please login to submit your rating.

Abstract 

Dataset I mainly consists of 30 subjects, which are respectively composed of gait data collected by mobile phone placed on arm, wrist, hand, waist, and ankle. This dataset is used to verify the impact of the mobile phone's placement on the recognition effect. Dataset II and Dataset III are composed of 113 subjects. Dataset II is the data collected from a mobile phone placed in the hand position, while Dataset III is the gait data collected from a mobile phone placed in the waist position. These two data sets are used primarily to verify the identification effect of the proposed model.

Instructions: 

There are five subfiles in the folder "DataSet I": arm, wrist, hand, waist, and ankle. Each subfile contains training set and test set files, and there are 30 .csv files under each training set and test set belonging to 30 subjects respectively.

The DataSet II and DataSet III folders directly contain two subfiles, the training set, and the test set, because they only collect the gait data of a certain location. There are 113 .csv files in both the training set and test set folders, which belonging to 113 subjects respectively.

       All data sets are raw data that has been stripped of the head and tail useless data.