Regresar

SMCS: Mobile Model Oriented to Cloud for the Automatic Classification of Environmental Sounds

Abstract:

This paper presents SMCS, a cloud-oriented mobile system model that uses a Convolutional Neural Network for the automatic classification of environmental sounds in real time. The model comprises an architectural schema with its corresponding deployment scheme in Google cloud services provider. Finally, the validation protocol of SMCS is applied in two experiments using respectively the base of free sounds FSDkaggle2018 and a selection of warning sounds extracted from the same sound base. The results of the validation of the model are promising with high values of precision in the classification of sounds, demonstrating that the SMCS model is expected to be a point of reference for the development of sound analysis systems, contributing to improving the quality of life of people with Hearing Impairment.