Events Detection for an Audio-Based Surveillance System

Abstract

The present research deals with audio events detection in noisy environments for a multimedia surveillance application. In surveillance or homeland security most of the systems aiming to automatically detect abnormal situations are only based on visual clues while, in some situations, it may be easier to detect a given event using the audio information. This is in particular the case for the class of sounds considered in this paper, sounds produced by gun shots. The automatic shot detection system presented is based on a novelty detection approach which offers a solution to detect abnormality (abnormal audio events) in continuous audio recordings of public places. We specifically focus on the robustness of the detection against variable and adverse conditions and the reduction of the false rejection rate which is particularly important in surveillance applications. In particular, we take advantage of potential similarity between the acoustic signatures of the different types of weapons by building a hierarchical classification system.

Low-Complexity Hybrid Time-Frequency Audio Signal Pattern Detection

“…In the following, we assume that the human voice (e.g., the Handel chorus or experimentally acquired voices) is the signal of interest, whereas non-human voice audio signals (e.g., tank 4 Typically, in the VAD literature N high → low > N low → high [8].…”

“…In [3], the authors characterize the relevant spectral peaks of different audio patterns (for health care purposes) in order to perform the recognition task. In [4], an audio-based recognition system for gun shot detection is presented and its robustness against variable and adverse conditions is analyzed. Different time and frequency domain metrics for audio-based context recognition systems are analyzed in [5], comparing system performance with the accuracy of human listeners performing the same task.…”

Feature extraction based on the high-pass filtering of audio signals for Acoustic Event Classification

“…can give relevant cues about the human presence and activity in a certain scenario (for example, in an office room). This information could be used in different applications, mainly in those with perceptually aware interfaces such as smart-rooms (Temko and Nadeu, 2006), automotive applications (Muller et al, 2008), mobile robots working in diverse environments (Chu et al, 2006) or surveillance systems (Clavel et al, 2005). Additionally, acoustic event detection and classification systems, can be used as a pre-processing stage for Automatic Speech Recognition (ASR) in such a way that this kind of sounds can be removed prior to the recognition process increasing its robustness.…”

Improving Detection of Acoustic Signals by Means of a Time and Frequency Multiple Energy Detector

“…Some of the most interesting ones are the surveillance applications in which the signals recorded by a set of microphones are processed to extract as much information as possible of the environment [1], [2]. Other related works include acoustic event detection in order to determine the presence of sounds in real life scenarios [3].…”