Auditory Support for Situation Awareness in Video Surveillance
MetadataShow full item record
We introduce a parameter mapping sonification to support situational awareness of surveillance operators during their task of monitoring video data. The presented auditory display produces a continuous ambient soundscape reflecting the changes in video data. For this purpose, we use low-level computer vision techniques, such as optical-flow extraction and background subtraction, and rely on the capabilities of the human auditory system for high-level recognition. Special focus is put on the mapping between video features and sound parameters. We optimize this mapping to provide a good interpretability of the sound pattern, as well as an aesthetic non-obtrusive sonification: precision of the conveyed information, psychoacoustic capabilities of the auditory system, and aesthetical guidelines of sound design are considered by optimally balancing the mapping parameters using gradient descent. A user study evaluates the capabilities and limitations of the presented sonification, as well as its applicability to supporting situational awareness in surveillance scenarios.