On the Task of Classifying Sound Patterns in Transport

The paper considers methods of collecting audio information about traffic events. The analysis of existing approaches to the extraction of audio patterns is given, the advantages and disadvantages are highlighted. The design of a neural network algorithm is performed. The requirements to the functionality of the data processing system are formed. A comparative analysis of the developed algorithm and examined its applicability to the task of classifying sound patterns in transportation. Ideas for further research to improve the performance of classification algorithms are formed.