Quartered Chirp Spectral Envelope for Whispered vs Normal Speech Classification

TL;DR

This paper introduces the quartered chirp spectral envelope, a novel feature for classifying whispered versus normal speech, which outperforms existing methods especially under noisy conditions.

Contribution

The paper proposes a new spectral feature combining chirp spectrum and quartered spectral envelope, improving whispered vs normal speech classification robustness.

Findings

Outperforms state-of-the-art in white noise conditions

Effective feature for whispered vs normal speech classification

Robustness demonstrated with CNN-based classifier

Abstract

Whispered speech as an acceptable form of human-computer interaction is gaining traction. Systems that address multiple modes of speech require a robust front-end speech classifier. Performance of whispered vs normal speech classification drops in the presence of additive white Gaussian noise, since normal speech takes on some of the characteristics of whispered speech. In this work, we propose a new feature named the quartered chirp spectral envelope, a combination of the chirp spectrum and the quartered spectral envelope, to classify whispered and normal speech. The chirp spectrum can be fine-tuned to obtain customized features for a given task, and the quartered spectral envelope has been proven to work especially well for the current task. The feature is trained on a one dimensional convolutional neural network, that captures the trends in the spectral envelope. The proposed system performs better than the state of the art, in the presence of white noise.