Multimodal Semantics Extraction from User-Generated Videos

Figure 2

Processing steps for extracting semantics from content data and context data (i.e., auxiliary sensor data captured by the electronic compass, the accelerometer, and the GPS receiver) captured by multiple cameras (denoted in the figure as “Cam”) in a public event.