Content-based Representations of audio using Siamese neural networks

TL;DR

This paper introduces a novel content-based audio retrieval method using Siamese neural networks to encode audio into vector representations, enabling effective retrieval of semantically similar recordings.

Contribution

The paper proposes a new approach that encodes audio into vectors with Siamese neural networks for improved content-based retrieval.

Findings

Effective retrieval of semantically similar audio using simple similarity measures

Siamese neural networks produce meaningful audio embeddings

Improved accuracy over traditional retrieval methods

Abstract

In this paper, we focus on the problem of content-based retrieval for audio, which aims to retrieve all semantically similar audio recordings for a given audio clip query. This problem is similar to the problem of query by example of audio, which aims to retrieve media samples from a database, which are similar to the user-provided example. We propose a novel approach which encodes the audio into a vector representation using Siamese Neural Networks. The goal is to obtain an encoding similar for files belonging to the same audio class, thus allowing retrieval of semantically similar audio. Using simple similarity measures such as those based on simple euclidean distance and cosine similarity we show that these representations can be very effectively used for retrieving recordings similar in audio content.