KU-HAR: An Open Dataset for Human Activity Recognition
Description
Human Activity Recognition (HAR) refers to the capacity of machines to perceive human actions. This dataset contains information on 18 different activities collected from 90 participants (75 male and 15 female) using smartphone sensors (Accelerometer and Gyroscope). It has 1945 raw activity samples collected directly from the participants, and 9185 subsamples extracted from them. The activities are: Stand➞ Standing still (1 min) Sit➞ Sitting still (1 min) Talk-sit➞ Talking with hand movements while sitting (1 min) Talk-stand➞ Talking with hand movements while standing or walking(1 min) Stand-sit➞ Repeatedly standing up and sitting down (5 times) Lay➞ Laying still (1 min) Lay-stand➞ Repeatedly standing up and laying down (5 times) Pick➞ Picking up an object from the floor (10 times) Jump➞ Jumping repeatedly (10 times) Push-up➞ Performing full push-ups (5 times) Sit-up➞ Performing sit-ups (5 times) Walk➞ Walking 20 meters (≈12 s) Walk-backward➞ Walking backward for 20 meters (≈20 s) Walk-circle➞ Walking along a circular path (≈ 20 s) Run➞ Running 20 meters (≈7 s) Stair-up➞ Ascending on a set of stairs (≈1 min) Stair-down➞ Descending from a set of stairs (≈50 s) Table-tennis➞ Playing table tennis (1 min) Contents of the attached .zip files are: 1.Raw_time_domian_data.zip➞ Originally collected 1945 time-domain samples in separate .csv files. The arrangement of information in each .csv file is: Column 1, 5➞ exact time (elapsed since the start) when the Accelerometer & Gyro output was recorded (in ms) Col. 2, 3, 4➞ Acceleration along X,Y,Z axes (in m/s^2) Col. 6, 7, 8➞ Rate of rotation around X,Y,Z axes (in rad/s) 2.Trimmed_raw_data.zip➞ Samples of the previous file after certain parts of the signals that contained no information on the corresponding activity were trimmed. 3.Time_domain_subsamples.zip➞ 9185 subsamples extracted from the 1945 collected samples in a single .csv file. Arrangement of information: Col. 1–1500, 1501–3000, 3001–4500➞ Acc.meter X, Y, Z axes readings Col. 4501–6000, 6001–7500, 7501–9000➞ Gyro X, Y, Z axes readings Col. 9001➞ Class ID (0 to 17, in the order mentioned above) Col. 9002➞ length of the subsample (each signal begins from the starting column and runs its course, the remaining columns are padded with zeros) Col. 9003➞ serial no. of the subsample 4.Frequency_features.zip➞ The 1500-point DFT output of each signal of 9185 subsamples in a single .csv file. The arrangement of information is the same as above. Samples were collected at 100 Hz, gravity acceleration was omitted from the Acc.meter data, and no filter was applied to remove noise. The dataset is free to download, modify, and use. More information is provided in the data paper which is currently submitted: N. Sikder, A.-A. Nahid, KU-HAR: An open dataset for heterogeneous human activity recognition, Pattern Recognit. Lett. (submitted). A preprint will be available soon. Backup: drive(dot)google.com/drive/folders/1cS29tHwlu9MGM9OYq8ApInz0eDnuyPFs