Additional Shared Decoder on Siamese Multi-view Encoders for Learning Acoustic Word Embeddings

TL;DR

This paper introduces a novel shared decoder architecture for Siamese multi-view encoders to improve acoustic word embeddings, achieving significant performance gains in word discrimination tasks.

Contribution

It proposes a shared decoder network integrated with Siamese multi-view encoders, enhancing the relationship between acoustic and text embeddings for better discrimination.

Findings

11.1% relative improvement in average precision on WSJ dataset

Effective in cross-view word discrimination tasks

Improved performance in word-level speech recognition

Abstract

Acoustic word embeddings --- fixed-dimensional vector representations of arbitrary-length words --- have attracted increasing interest in query-by-example spoken term detection. Recently, on the fact that the orthography of text labels partly reflects the phonetic similarity between the words' pronunciation, a multi-view approach has been introduced that jointly learns acoustic and text embeddings. It showed that it is possible to learn discriminative embeddings by designing the objective which takes text labels as well as word segments. In this paper, we propose a network architecture that expands the multi-view approach by combining the Siamese multi-view encoders with a shared decoder network to maximize the effect of the relationship between acoustic and text embeddings in embedding space. Discriminatively trained with multi-view triplet loss and decoding loss, our proposed approach achieves better performance on acoustic word discrimination task with the WSJ dataset, resulting in 11.1% relative improvement in average precision. We also present experimental results on cross-view word discrimination and word level speech recognition tasks.