Automatic Analysis of the Emotional Content of Speech in Daylong Child-Centered Recordings from a Neonatal Intensive Care Unit

TL;DR

This paper introduces a large-scale dataset of daylong neonatal ICU recordings and develops an automatic speech emotion recognition system to analyze emotional content, demonstrating effective domain adaptation and active learning techniques.

Contribution

The study presents the first large-scale, real-world neonatal ICU audio dataset with emotion labels and a novel SER system tailored for this domain, including evaluation of domain adaptation methods.

Findings

Achieved 73.4% UAR in emotion classification

Active learning outperformed other domain adaptation methods

Demonstrated feasibility of automatic emotion analysis in neonatal ICU audio

Abstract

Researchers have recently started to study how the emotional speech heard by young infants can affect their developmental outcomes. As a part of this research, hundreds of hours of daylong recordings from preterm infants' audio environments were collected from two hospitals in Finland and Estonia in the context of so-called APPLE study. In order to analyze the emotional content of speech in such a massive dataset, an automatic speech emotion recognition (SER) system is required. However, there are no emotion labels or existing indomain SER systems to be used for this purpose. In this paper, we introduce this initially unannotated large-scale real-world audio dataset and describe the development of a functional SER system for the Finnish subset of the data. We explore the effectiveness of alternative state-of-the-art techniques to deploy a SER system to a new domain, comparing cross-corpus generalization, WGAN-based domain adaptation, and active learning in the task. As a result, we show that the best-performing models are able to achieve a classification performance of 73.4% unweighted average recall (UAR) and 73.2% UAR for a binary classification for valence and arousal, respectively. The results also show that active learning achieves the most consistent performance compared to the two alternatives.