An Iterative Deep Learning Framework for Unsupervised Discovery of Speech Features and Linguistic Units with Applications on Spoken Term Detection

TL;DR

This paper introduces an iterative deep learning framework called MAT-DNN that automatically discovers speech features and linguistic units from unlabeled speech data, improving spoken term detection performance in a zero-resource setting.

Contribution

The paper presents a novel iterative framework combining a Multi-layered Acoustic Tokenizer and Deep Neural Network for unsupervised speech feature and unit discovery.

Findings

Improved speech feature quality for zero-resource tasks

Enhanced acoustic token discovery through iterative feedback

Better spoken term detection performance

Abstract

In this work we aim to discover high quality speech features and linguistic units directly from unlabeled speech data in a zero resource scenario. The results are evaluated using the metrics and corpora proposed in the Zero Resource Speech Challenge organized at Interspeech 2015. A Multi-layered Acoustic Tokenizer (MAT) was proposed for automatic discovery of multiple sets of acoustic tokens from the given corpus. Each acoustic token set is specified by a set of hyperparameters that describe the model configuration. These sets of acoustic tokens carry different characteristics fof the given corpus and the language behind, thus can be mutually reinforced. The multiple sets of token labels are then used as the targets of a Multi-target Deep Neural Network (MDNN) trained on low-level acoustic features. Bottleneck features extracted from the MDNN are then used as the feedback input to the MAT and the MDNN itself in the next iteration. We call this iterative deep learning framework the Multi-layered Acoustic Tokenizing Deep Neural Network (MAT-DNN), which generates both high quality speech features for the Track 1 of the Challenge and acoustic tokens for the Track 2 of the Challenge. In addition, we performed extra experiments on the same corpora on the application of query-by-example spoken term detection. The experimental results showed the iterative deep learning framework of MAT-DNN improved the detection performance due to better underlying speech features and acoustic tokens.