Deep learning for human activity recognition: A resource efficient implementation on low-power devices

Keyphrases

• Deep Learning 100%
• Human Activity Recognition 100%
• Efficient Implementation 100%
• Low-power Devices 100%
• Feature Generation 66%
• Sensor Node 66%
• Recognition Method 33%
• Healthcare Applications 33%
• Sensor Orientation 33%
• Sensor Data Collection 33%
• Inertial Sensors 33%
• State-of-the-art Techniques 33%
• Real-world Activity 33%
• Well-being 33%
• Specific Applications 33%
• Temporal Convolutional Network 33%
• Sensor Data 33%
• Activity Datasets 33%
• Mobile Sensors 33%
• Mobile Devices 33%
• Invariance 33%
• Orientation Sensor 33%
• Machine Learning Approach 33%
• Inertial Data 33%
• Computation Time 33%
• Activity Recognition 33%
• Sensor Placement 33%
• Real-time Classification 33%
• Low-cost Wearable 33%
• Generation Process 33%
• Contextual Information 33%
• Acquisition Rate 33%
• Spectral Domain 33%
• Learning Methodology 33%
• Sports Applications 33%
• Offline Processing 33%

Computer Science

• Deep Learning Method 100%
• Human Activity Recognition 100%
• Efficient Implementation 100%
• Sensor Node 66%
• Feature Generation 66%
• Sensor Placement 33%
• Contextual Information 33%
• Systematic Analysis 33%
• Computation Time 33%
• Activity Recognition 33%
• Mobile Device 33%
• Spectral Domain 33%
• Mobile Sensor 33%
• Machine Learning Approach 33%
• Wearable Device 33%