MultiTask-LSTM-HAR

This is the implementation of the paper "Multitask LSTM Model for Human Activity Recognition and Intensity Estimation Using Wearable Sensor Data" (IEEE IoT Journal), 2020

Dataset

The dataset splits can be downloaded here, please create data/ folder, download the csv files and place them in the data/ folder.

Requirements

• python 3.5.8 or higher
• matplotlib
• tensorflow-gpu 1.6.0 or higher

Create and setup the virtual Environment

python3 -m venv ./env

source ./env/bin/activate

pip install -r requirements.txt

Training and Testing

When the virtual environment is activated:

python3 wearable-main.py --function clfOnly --saveModel --testSubject 10

Setting the hyperparameters is optional. For details:

python3 wearable-main.py --help

References

If this repository was useful for your research, please cite.

@ARTICLE{UML-HAR20,
  author={Barut, Onur and Zhou, Li and Luo, Yan},
  journal={IEEE Internet of Things Journal}, 
  title={Multitask LSTM Model for Human Activity Recognition and Intensity Estimation Using Wearable Sensor Data}, 
  year={2020},
  volume={7},
  number={9},
  pages={8760-8768},
  doi={10.1109/JIOT.2020.2996578}}