Research Information System
Multimedia Tools and Applications, vol.81, no.6, pp.7969-7991, 2022 (SCI-Expanded)
© 2022, The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature.Modern video surveillance systems consist of a network of many video cameras. Constantly video camera systems are being installed for security reasons in prisons, elevators, automatic teller machines and more. Usually, video cameras are connected to a display screen from which security personnel monitor suspicious activity. As security personnel monitor multiple locations simultaneously, this manual task is labor intensive and inefficient. These camera systems have some other drawbacks such that they have limited coverage and security personnel cannot see all the points even though they are looking at the camera. Therefore, most of the time, some other sensors should accompany to video cameras. Although audio surveillance is in its early stage, there has been considerable amount of work in this area in the last decade. On the other hand, currently, there are no practical audio surveillance solutions for security on the market. In this paper, audio surveillance is integrated to current video surveillance systems using deep learning. We develop a complete system and show a working prototype. It is encouraging to see that the system is good enough and can be used in real life.