Center for Machine Learning and Intelligent Systems 		About  Citation Policy  Donate a Data Set  Contact


Repository Web            Google
View ALL Data Sets

× Check out the beta version of the new UCI Machine Learning Repository we are currently testing! Contact us if you have any issues, questions, or concerns. Click here to try out the new site.

WISDM Smartphone and Smartwatch Activity and Biometrics Dataset Data Set
Download: Data Folder, Data Set Description

Abstract: Contains accelerometer and gyroscope time-series sensor data collected from a smartphone and smartwatch as 51 test subjects perform 18 activities for 3 minutes each.

Data Set Characteristics:  

Multivariate, Time-Series

Number of Instances:

15630426

Area:

Computer

Attribute Characteristics:

Real

Number of Attributes:

6

Date Donated

2019-10-06

Associated Tasks:

Classification

Missing Values?

N/A

Number of Web Hits:

70644

Source:

Dr. Gary Weiss, gaweiss '@' fordham.edu, Computer and Information Sciences Department, Fordham University.


Data Set Information:

For a detailed description of the dataset, please see the following pdf file that is stored with the data: WISDM-dataset-description.pdf. The raw accelerometer and gyroscope sensor data is collected from the smartphone and smartwatch at a rate of 20Hz. It is collected from 51 test subjects as they perform 18 activities for 3 minutes apiece. The sensor data for each device (phone, watch) and type of sensor (accelerometer, gyroscope) is stored in a different directory (so there are 4 data directories). In each directory there are 51 files corresponding to the 51 test subjects. The format of every entry is the same: . The descriptions of these attributes are provided with the attribute information. In addition to the raw time-series sensor data we also generate examples that describe the sensor data using a 10-second window. See the dataset description document for details. Although this data can most naturally be used for activity recognition, it can also be used to build behavioral biometric models since each sensor reading is associated with a specific subject.


Attribute Information:

subject-id: value from 1600- 1650 that identifies one of the 51 test subjects
activity-code: character between 'A' and 'S' (no 'N') that identifies the activity. The mapping from code to activity is provided in the activity_key.txt file and in our dataset description document.
timestamp: Unix time (integer)
x: represents the sensor reading (accelerometer or gyroscope) for the x dimension
y: represents the sensor reading (accelerometer or gyroscope) for the y dimension
z: represents the sensor reading (accelerometer or gyroscope) for the z dimension


Relevant Papers:

Gary M. Weiss, Kenichi Yoneda, and Thaier Hayajneh. Smartphone and Smartwatch-Based Biometrics Using Activities of Daily Living. IEEE Access, 7:133190-133202, Sept. 2019.



Citation Request:

Please cite the IEEE Access article: Smartphone and Smartwatch-Based Biometrics Using Activities of Daily Living. IEEE Access, 7:133190-133202, Sept. 2019.

Supported By:

 In Collaboration With:
About  ||  Citation Policy  ||  Donation Policy  ||  Contact  ||  CML