Research‎ > ‎

Situation Awareness

Understanding current situation is important for Human-Robot Interaction, as appropriate perception generates proper action for robot.
In this research, environmental sound is used as input for situation awareness system. From noisy signal, target sound is extracted and is classified into pre-defined class.

  • Blind Source Separation and Blind Source Extraction for noisy audio signal
  • Emergency verification using the signal analysis
  • Emergency classification using machine learning with big data