Introduction

This code is based on the paper entitled "An Energy-based Method for Orientation Correction of EMG Bracelet Sensors in Hand Gesture Recognition Systems" by Laboratorio de Investigación en Inteligencia y Visión Artificial “Alan Turing” Escuela Politecnica Nacional (C) Copyright 2020

Data-set can be found in the following link:

https://laboratorio-ia.epn.edu.ec/es/recursos/dataset/2020_emg_dataset_612

Data conversion from .json to .mat

• Download the Training users dataset and paste it into the folder "trainingJSON".

• Download the Testing users dataset and paste it into the folder "testingJSON".

Run jsontomat.m

• After executing jsontomat.m two folders will be generated:

  • TrainingData
  • TestingData

• Copy the user folders generated in "TrainingData"(*.mat format) and paste them in the "Data\General\training" folder.

• Copy the user folders generated in "TestingData"(*.mat format) and paste them in the "Data\General\testing" and "Data\Specific" folders.

Instructions for Matlab:

Run Main.m

Select the experiment to run

Parameters:

• Syncro: number of synchronization signals to test (1-4).

Menu:

• Gesture Recognition

  • [1] Experiment 1 : This option runs the experiment 1

  • [2] Experiment 2 : This option runs the experiment 2

  • [3] Experiment 3 : This option runs the experiment 3

  • [4] Experiment 4 : This option runs the experiment 4

  • [5] Exit

    Select an option to run:

    • Classification Models
      • [1] General : This option runs the general model in the current experiment
      • [2] Especific : This option runs the specific model in the current experiment
      • [3] Exit

Evaluation

The result obtained in json format must be evaluated in the following link:

https://aplicaciones-ia.epn.edu.ec/webapps/home/session.html?app=EMG%20Gesture%20Recognition%20Evaluator