Pawel Zwan

Dangerous Sound Event Recognition Using Support Vector Machine Classifiers

A method of recognizing events connected to danger based on their acoustic representation through Support Vector Machine classification is presented. The method proposed is particularly useful in an automatic surveillance system. The set of 28 parameters used in the classifier consists of dedicated parameters and MPEG-7 features. Methods for parameter calculation are presented, as well as a design of SVM model used for classification. The performance of the classifier was tested on a set of 372 example sounds, yielding high accuracy.

Report of the ISMIS 2011 contest: music information retrieval

This report presents an overview of the data mining contest organized in conjunction with the 19th International Symposium on Methodologies for Intelligent Systems (ISMIS 2011), in days between Jan 10 and Mar 21, 2011, on TunedIT competition platform. The contest consisted of two independent tasks, both related to music information retrieval: recognition of music genres and recognition of instruments, for a given music sample represented by a number of pre-extracted features. In this report, we describe aim of the contest, tasks formulation, procedures of data generation and parametrization, as well as final results of the competition.

Automatic assessment of the motor state of the Parkinson's disease patient-a case study

This paper presents a novel methodology in which the Unified Parkinsons Disease Rating Scale (UPDRS) data processed with a rule-based decision algorithm is used to predict the state of the Parkinsons Disease patients. The research was carried out to investigate whether the advancement of the Parkinsons Disease can be automatically assessed. For this purpose, past and current UPDRS data from 47 subjects were examined. The results show that, among other classifiers, the rough set-based decision algorithm turned out to be most suitable for such automatic assessment.Virtual slidesThe virtual slide(s) for this article can be found here:http://www.diagnosticpathology.diagnomx.eu/vs/1563339375633634.

Verification of the Parameterization Methods in the Context of Automatic Recognition of Sounds Related to Danger

Digital signal processing of sound is a domain with numerous applications in the telecommunications and informatics. These well-developed algorithms of the analysis of sound can be also applied in the field of security systems, where traditional monitoring is still based mainly on video cameras. The commonly used monitoring cameras can be equipped with additional microphones and the audio content can be analyzed by a monitoring program running on a dedicated hardware. This application can automatically detect in the audio stream events like a broken window, gunshot, explosion, or scream. One of the main parts of this system is a parameterization block. In this article two parameterization methods are proposed for this purpose. The first is based on the frequency analysis of the examples of the sound events. The second is based on using a standardized set of audio MPEG-7 and cepstral descriptors. The feature vectors calculated by these two methods have been used for the training of two intelligent classifiers: a support vector machines classifier (SVM) and a neural networks perceptron (NNP). The classifiers have been verified using of the cross-validation method. The results have been compared and conclusions derived. The application of the results in a system working in real conditions is presented and discussed at the end of the article. The work has been done in the frame of the international project “INDECT” (Intelligent Information System Supporting Observation, Searching and Detection for Security of Citizens in Urban Environment).

Wavelet‐based automatic recognition of musical instruments

The objective of the present work is to automatically extract information from monophonic sounds. This process consists of several stages, namely, preprocessing, parameterization, and classification. This paper shows a thorough study on the wavelet‐based parameterization of musical instrument sounds and automatic recognition by means of artificial neural networks (ANNs). First, an engineered method of pitch detection is presented and exemplified by several analyses. A short discussion on error associated with automatic pitch tracking is also included. Then, examples of time‐frequency analyses of various musical instrument groups are presented. The analyses are performed employing a database containing musical sounds recorded at the Sound and Vision Engineering Department, Technical University of Gdansk. On the basis of such analyses a set of parameters is derived. Feature vector properties are then discussed. For that purpose Fisher statistics is used. It allows checking the separability between musical i...