Walk to Show Your Identity: Gait-based Seamless User Authentication Framework Using Deep Neural Network

This is python code for the paper work published in WearSys '19: The 5th ACM Workshop on Wearable Systems and Applications. You can access to the paper through this link

Prerequisities

• Language: Python
• Required Packages: numpy, pandas, matplotlib, scipy, sklearn, tensorflow
• To install the required package, type the following command (To install Tensorflow, visit the official Tensorflow Webpage link):

1. Python 2

pip install numpy pandas matplotlib scipy sklearn

2. Python 3

pip3 install numpy pandas matplotlib scipy sklearn

Running the code

1. Data Filtering & Visualization of Frequency Distribution

python3 data_filter_fft.py
python3 valid_start_end.py

2. Gait Cycle Detection: slice walk cycles from the data sequences

python3 cycle_detection.py

3. Interpolation: make walk cycles consistent in length

python3 interpolation.py

4. Cycle Filtering: filter out noisy cycles

python3 cycle_filter.py

5. Classification: DNN (Multi Layer Perceptron), CNN, and RNN (LSTM)

python3 DNN.py
python3 CNN.py
python3 RNN.py