Classificação de Cenários Acústicos em Dispositivos Auditivos: Uma Revisão Sistemática da Literatura sobre Desafios, Avanços e Perspectivas

Acoustic Scene Classification (ASC) is an area of growing relevance,with applications ranging from assistive devices, such as hearingaids, to advanced wearable technologies (hearables). This paperpresents a Systematic Literature Review (SLR) that analyzes themain adaptive and machine learning-based methods used in ASC,with a focus on hearing devices. The challenges related to computationalresource limitations, energy consumption and real-timeoperation, especially in dynamic environments, are discussed. Thereview highlights recent advances, such as the use of generativeprobabilistic models and convolutional neural networks, as well ashybrid approaches that combine cloud computing and edge computingfor greater efficiency. The results show that, despite significantprogress, there are still important technical barriers, such as theneed for more efficient, customizable and robust algorithms to operatein real conditions. This study contributes by identifying gapsin the literature and suggesting future directions to improve theintegration of ASC in hearing devices.