Human Activity Recognition from different datasets

Human Activity recognition using 1D Convolutional Neural Network for different datasets

Dataset

• HASC
• WISDM
• SinlgeChest

Tools

• Jupyter Notebook

1D CNN Model

model = Sequential()
model.add(Conv1D(filters=64, kernel_size=3, activation='relu', input_shape=(n_timesteps,n_features)))
model.add(Conv1D(filters=64, kernel_size=3, activation='relu', padding = 'same'))
model.add(Dropout(0.4))
model.add(MaxPooling1D(pool_size=2))
model.add(Flatten())
model.add(Dense(100, activation='relu'))
model.add(Dense(n_outputs, activation='softmax'))

Description

• Activity Types for HASC dataset
  

• Activity plot for WISDM dataset
  

• Activities for SingleChest dataset
  

• Confusion Matrix using CNN for HASC dataset
  

• Confusion Matrix using CNN for WISDM dataset
  

• Confusion Matrix using CNN for SingleChest dataset