https://github.com/codeamt/human-activity-recognition

Human Activity Recognition (HAR) with Samsung Mobile Sensor Data and Fast v2
https://github.com/codeamt/human-activity-recognition

Last synced: about 2 months ago
JSON representation

Human Activity Recognition (HAR) with Samsung Mobile Sensor Data and Fast v2

• Host: GitHub
• URL: https://github.com/codeamt/human-activity-recognition
• Owner: codeamt
• Created: 2020-12-18T08:56:57.000Z (about 4 years ago)
• Default Branch: master
• Last Pushed: 2022-07-10T00:08:06.000Z (over 2 years ago)
• Last Synced: 2023-10-20T04:52:17.526Z (about 1 year ago)
• Language: Jupyter Notebook
• Size: 98.6 KB
• Stars: 0
• Watchers: 2
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Human Activity Recognition

 Human Activity Recognition (HAR) with Samsung Mobile Sensor Data and Fast v2

 

## Dataset

[1] Records 30 volunteer subjects, ages 19-48 years, performed six activities with mobile embedded accelerometer and gyroscope around the waist. captured 3-axial linear acceleration and 3-axial angular velocity at a constant rate of 50Hz. The obtained dataset has been randomly partitioned into two sets, where 70% of the volunteers were selected for generating the training data and 30% for the test data.

Classes:

0: Walking

1: Walking Upstairs

2: Walking Downstairs

3: Sitting

4: Standing

5: Laying

Training Examples: 7352

Test Examples: 2947

[1] Davide Anguita, Alessandro Ghio, Luca Oneto, Xavier Parra and Jorge L. Reyes-Ortiz. 

A Public Domain Dataset for Human Activity Recognition Using Smartphones. 

21th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, ESANN 2013. Bruges, Belgium 24-26 April 2013.

## Training

HAR Model Architecture developed with PyTorch, with 

a series on convolutional, dropout, and adaptive pooling layers and trained with the Fast.ai (v2) library:



## Inference

88% Accuracy