A Semi-spontaneous Dutch Speech Dataset for Speech Enhancement and Speech Recognition

TL;DR

This paper introduces DRES, a semi-spontaneous Dutch speech dataset recorded in noisy indoor environments, and evaluates the impact of speech enhancement on recognition performance using state-of-the-art models.

Contribution

The creation of DRES, a realistic Dutch speech dataset, and the comprehensive evaluation of speech enhancement and recognition models in real-world noisy conditions.

Findings

Five ASR models achieved WERs below 22% on DRES.

Modern single-channel speech enhancement did not improve ASR performance in realistic scenarios.

Evaluation in real-world conditions is crucial for assessing speech processing models.

Abstract

We present DRES: a 1.5-hour Dutch realistic elicited (semi-spontaneous) speech dataset from 80 speakers recorded in noisy, public indoor environments. DRES was designed as a test set for the evaluation of state-of-the-art (SOTA) automatic speech recognition (ASR) and speech enhancement (SE) models in a real-world scenario: a person speaking in a public indoor space with background talkers and noise. The speech was recorded with a four-channel linear microphone array. In this work we evaluate the speech quality of five well-known single-channel SE algorithms and the recognition performance of eight SOTA off-the-shelf ASR models before and after applying SE on the speech of DRES. We found that five out of the eight ASR models have WERs lower than 22% on DRES, despite the challenging conditions. In contrast to recent work, we did not find a positive effect of modern single-channel SE on ASR performance, emphasizing the importance of evaluating in realistic conditions.