Activity Recognition
HAR with Smartphones
Using a smartphone to recognize human activities
classification har human activity recognition smartphone accelerometer walking

The dataset was created by recording 30 participants carrying smartphone on their waist.

Using the accelerometer and gyroscope, they captured linear acceleration and angular velocity at a rate of 50Hz.

We were able to classify the following activities:

  • Walking
  • Walking upstairs
  • Walking downstairs
  • Sitting
  • Standing

Such a dataset could have applications in elderly care, health, or occupational safety. Especially in areas where no cameras can be used.

Results from deployment:
96%
1.5kB
Accuracy
RAM used