Reducing Model Complexity for DNN Based Large-Scale Audio Classification

TL;DR

This paper investigates large-scale audio classification using AudioSet, comparing neural network models, and proposes strategies to reduce CNN model complexity with minimal performance loss.

Contribution

It introduces two methods for creating low-dimensional embeddings that significantly reduce CNN model parameters while maintaining accuracy.

Findings

CNN outperforms MLP and RNN in accuracy

Proposed strategies reduce CNN parameters to 1/22 with slight performance degradation

Effective model simplification for practical large-scale audio classification

Abstract

Audio classification is the task of identifying the sound categories that are associated with a given audio signal. This paper presents an investigation on large-scale audio classification based on the recently released AudioSet database. AudioSet comprises 2 millions of audio samples from YouTube, which are human-annotated with 527 sound category labels. Audio classification experiments with the balanced training set and the evaluation set of AudioSet are carried out by applying different types of neural network models. The classification performance and the model complexity of these models are compared and analyzed. While the CNN models show better performance than MLP and RNN, its model complexity is relatively high and undesirable for practical use. We propose two different strategies that aim at constructing low-dimensional embedding feature extractors and hence reducing the number of model parameters. It is shown that the simplified CNN model has only 1/22 model parameters of the original model, with only a slight degradation of performance.