We collected a database of human activities. The constructed dataset comprises of 18 activities plus Null activities.
Thirteen subjects (5 female and 8 male) aged between 19 and 45 are asked to worn e-Shoes and Samsung Gear S2 smart-watch on the preferred hand (see Fig. 1 for device setup).
The subjects are asked to sign the consent forms and given the list of 18 activities.
Before performing activities, subjects are asked to perform the "kick and hit-the-hand" activity to make highly distinctive signals for synchronizing sensors and video, and then resting for 10 seconds before performing activities. There are 9 in-door activities denoted by (i) and 9 sport out-door activities dennoted by (o) as listed in Fig. 2.
During performing the pre-defined activities, the subject could perform any arbitrary activity out of 18 activity list. We consider all of activities out of interest as Null activities. Duration time for each activity varies from 3 to 10 minutes. In addition, several surveillance cameras are installed in the kitchen, living room, and outdoor space to capture the activity videos which are used later for annotation. Two people have annotated the whole dataset using ELAN software. Figure 3 shows the interface of annotation and synchronization. Only signal corresponds to predefined activities are labeled and the other are marked as Null.

Fig 1. Device setup

Fig 2. List of 9 activities

Fig 3. Signal synchronization in 19NonSens dataset

19NonSense dataset.

The dataset is released for academic research only and is free to researchers from educational or research institutes for non-commercial purposes.
Link for dowdloading the dataset: [Drive]
Please read information in 19NonSense.pdf in the downloading folder for more information
Furthermore, these publications should cite the following papers:
Cuong Pham, Son Nguyen-Thai, Huy Tran-Quang, Son Tran, Hai Vu, Thanh-Hai Tran, Thi-Lan Le*, SensCapsNet: Deep Neural Network for Non-obtrusive Sensing based Human Activity Recognition, IEEE Access, 2020.
Home page