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