Data Augmentation of Room Classifiers using Generative Adversarial Networks

TL;DR

This paper introduces a GAN-based data augmentation technique for room classification using reverberant speech, significantly improving classifier accuracy without additional data collection.

Contribution

It proposes a novel acoustic environment representation and a GAN training method to generate realistic artificial room data for improved classification.

Findings

Test accuracy increased from 89.4% to 95.5%.

GAN-generated data effectively enhances classifier performance.

Method reduces need for extensive real-world data collection.

Abstract

The classification of acoustic environments allows for machines to better understand the auditory world around them. The use of deep learning in order to teach machines to discriminate between different rooms is a new area of research. Similarly to other learning tasks, this task suffers from the high-dimensionality and the limited availability of training data. Data augmentation methods have proven useful in addressing this issue in the tasks of sound event detection and scene classification. This paper proposes a method for data augmentation for the task of room classification from reverberant speech. Generative Adversarial Networks (GANs) are trained that generate artificial data as if they were measured in real rooms. This provides additional training examples to the classifiers without the need for any additional data collection, which is time-consuming and often impractical. A representation of acoustic environments is proposed, which is used to train the GANs. The representation is based on a sparse model for the early reflections, a stochastic model for the reverberant tail and a mixing mechanism between the two. In the experiments shown, the proposed data augmentation method increases the test accuracy of a CNN-RNN room classifier from 89.4% to 95.5%.