Layer 1 – Physical Signals
Layer 1 is the foundation of the conceptual model.
It deals with the raw data captured by physical sensors and their preprocessing into virtual sensors — representations ready for higher-level analysis.
Key Concepts
A virtual sensor is a combination of:
- One or more physical sensors (e.g., MoCap, accelerometer, Kinect, respiration band).
- Signal conditioning (denoising, filtering, synchronization).
- Feature-specific extraction (e.g., 3D joint trajectories, silhouettes, barycenter).
Examples of Physical & Virtual Sensors
| Virtual Sensor / Signal | Description | Implemented |
|---|---|---|
| Trajectories | Positional data (2D/3D positions of joints and barycenter) from MoCap, video, or RGB-D sensors (e.g., Kinect). | |
| Bounding Space / Convex Hull | Minimum polygon (2D) or volume (3D) surrounding a point cloud (MoCap) or a body silhouette. | |
| Accelerations | Measures from accelerometers and gyroscopes. | |
| Physiological Sensors | EMG, EEG, ECG, and related physiological data. | |
| Respiration | Signals from dedicated respiration sensors or microphones. | |
| Nonverbal Vocal Utterances | Short vocalizations linked to movement (e.g., kiai in Karate, dance utterances). | |
| Floor Feet Pressure | Weight distribution across feet, measured with a sensitive floor. |