Acoustic Classification System Research Articles

Interaction between annoyance, indoor noise levels and acoustic classification of buildings

Assessment of annoyance "at home" from environmental noises has been widely investigated so far and the ISO/TS 15666:2021 was developed to lead the socio-acoustic surveys. On the other hand, the rating of buildings' acoustical performance considering also the indoor noises, has been also well concerned in building acoustics and the studies have ended up with the ISO/TS 19488:2021 covering the acoustic classification system for buildings. Basically the rating system needs to be supported by the subjective tests in the field or in laboratories, to acquire data about the annoyance/disturbance or satisfaction of residents. If the target is to design the healthy, comfortable and sustainable acoustical environment, both technical standards might be harmonized in the future. In this paper, based on the dose/response relationships with respect to the indoor noise levels, an approach is proposed to translate the acoustic classes proposed in ISO/TS 19488, into the annoyance boundaries in terms of different scales (verbal/numerical and the HA% ) referred in ISO/TS 15666. The results from the previous laboratory and field studies conducted by the authors, have been used for verification of this approach.

UALF: A learnable front-end for intelligent underwater acoustic classification system

In practical ocean engineering application, variable target characteristics and inevitable environmental noise will decrease the recognition accuracy of underwater acoustic classification system. Thus, an intelligent Underwater Acoustic classification system with a Learnable Front-end (UALF) is proposed to improve automation. Similar to the cognitive process of the human, UALF extracts characteristics of underwater acoustic from multiple perspectives and adaptively adjusts the feature extraction and classifier modules to suit different application scenarios. Datasets under different recording conditions and scenarios are used for the experiments. The results showed that UALF outperforms all baseline methods, which verified its intelligence and feasibility for practical engineering application.

A multi-view CNN-based acoustic classification system for automatic animal species identification

Automatic identification of animal species by their vocalization is an important and challenging task. Although many kinds of audio monitoring system have been proposed in the literature, they suffer from several disadvantages such as non-trivial feature selection, accuracy degradation because of environmental noise or intensive local computation. In this paper, we propose a deep learning based acoustic classification framework for Wireless Acoustic Sensor Network (WASN). The proposed framework is based on cloud architecture which relaxes the computational burden on the wireless sensor node. To improve the recognition accuracy, we design a multi-view Convolution Neural Network (CNN) to extract the short-, middle-, and long-term dependencies in parallel. The evaluation on two real datasets shows that the proposed architecture can achieve high accuracy and outperforms traditional classification systems significantly when the environmental noise dominate the audio signal (low SNR). Moreover, we implement and deploy the proposed system on a testbed and analyse the system performance in real-world environments. Both simulation and real-world evaluation demonstrate the accuracy and robustness of the proposed acoustic classification system in distinguishing species of animals.

Bee Swarm Activity Acoustic Classification for an IoT-Based Farm Service.

Beekeeping is one of the widespread and traditional fields in agriculture, where Internet of Things (IoT)-based solutions and machine learning approaches can ease and improve beehive management significantly. A particularly important activity is bee swarming. A beehive monitoring system can be applied for digital farming to alert the user via a service about the beginning of swarming, which requires a response. An IoT-based bee activity acoustic classification system is proposed in this paper. The audio data needed for acoustic training was collected from the Open Source Beehives Project. The input audio signal was converted into feature vectors, using the Mel-Frequency Cepstral Coefficients (with cepstral mean normalization) and Linear Predictive Coding. The influence of the acoustic background noise and denoising procedure was evaluated in an additional step. Different Hidden Markov Models’ and Gaussian Mixture Models’ topologies were developed for acoustic modeling, with the objective being to determine the most suitable one for the proposed IoT-based solution. The evaluation was carried out with a separate test set, in order to successfully classify sound between the normal and swarming conditions in a beehive. The evaluation results showed that good acoustic classification performance can be achieved with the proposed system.

REVIEW ON VEHICULAR SPEED, DENSITY ESTIMATION AND CLASSIFICATION USING ACOUSTIC SIGNAL

Traffic monitoring and parameters estimation from urban to non urban (battlefield environment) traffic is fast-emerging field based on acoustic signals. We present here a comprehensive review of the state-of-the-art acoustic signal for vehicular speed estimation, density estimation and classification, critical analysis and an outlook to future research directions. This field is of increasing relevance for intelligent transport systems (ITSs). In recent years video monitoring and surveillance systems has been widely used in traffic management and hence traffic parameters can be achieved using such systems, but installation, operational and maintenance cost associated with these approaches are relatively high compared to the use of acoustic signal which is having very low installation and maintenance cost. The classification process includes sensing unit, class definition, feature extraction, classifier application and system evaluation. The acoustic classification system is part of a multi sensor real time environment for traffic surveillance and monitoring. Classification accuracy achieved by various studied algorithms shows very good performance for the 'Heavy Weight' class of vehicles as compared to the other category Light Weight. Also a slight performance degrades as vehicle speed increases. Vehicular speed estimation corresponds to average speed and traffic density measurement, and can be substantially used for traffic signal timings optimization.

Bed diversity in the shallow water environment of Pappas lagoon in Greece

Acoustic classification systems and the Sediment Trend Analysis method were used to identify and map the bed diversity in a very shallow (<3.0m), coastal lagoon, Pappas lagoon, Western Greece. Analogue acoustic data, collected by means of a 100 kHz side scan sonar system, were digitized and classified into six acoustic classes using recently developed acoustic classification systems (SonarClass, TargAn). By comparing the acoustic classes to ground truth data consisting of sediment grain size and visual inspection of the lagoon-bed, it is demonstrated that the six acoustic classes correlate well with the predominant surface sediment types and vegetation. Thus the spatial distribution of the classes can be considered to represent the spatial pattern of the sedimentary assemblages of the lagoon. The grain size trend analysis identified three dominant sediment pathways and directional trends which could be related to the predominant wind direction, the sediment influx through the inlets and the sediment supply from a small stream in the southern part of the lagoon. The integration of acoustic and sedimentological data together with advanced data processing systems leads to a better understanding of the sedimentary, morphological and biological processes in a shallow lagoon in different spatial and temporal scales and will therefore be beneficial to both sedimentological and biotic-diversity studies.

Acoustic classification using fuzzy sets

A prototype acoustic classification system is designed using fuzzy sets. Parameters relevant to classification are extracted from acoustic signals. Given n parameters, an n-dimensional hypercube is constructed. Each contact is represented by a vector within this hypercube. Known contact classifications map along the edges of the hypercube. Unknown contacts are represented by vectors whose elements are multivalent. The distances between the unknown contact and the known contact classifications are calculated. The contact classification is determined by a radius of similarity which is computed as one-half of the average distance between all known contacts.