Ensemble Of Deep Neural Networks For Acoustic Scene Classification

TL;DR

This paper explores the adaptation of state-of-the-art deep neural networks, originally designed for image classification, to acoustic scene classification, demonstrating performance improvements through ensemble methods.

Contribution

It introduces modified DNN architectures for acoustic scene classification and empirically evaluates their effectiveness on a standard dataset, achieving notable performance gains.

Findings

Best model improves baseline score by 3.1% on test set

Achieves 10% improvement on development set

Ensemble of DNNs enhances classification accuracy

Abstract

Deep neural networks (DNNs) have recently achieved great success in a multitude of classification tasks. Ensembles of DNNs have been shown to improve the performance. In this paper, we explore the recent state-of-the-art DNNs used for image classification. We modified these DNNs and applied them to the task of acoustic scene classification. We conducted a number of experiments on the TUT Acoustic Scenes 2017 dataset to empirically compare these methods. Finally, we show that the best model improves the baseline score for DCASE-2017 Task 1 by 3.1% in the test set and by 10% in the development set.