Multitask vocal burst modeling with ResNets and pre-trained paralinguistic Conformers

TL;DR

This paper explores various modeling approaches for vocal burst analysis, demonstrating significant improvements with image classification models and investigating the potential of pre-trained Conformer models, while questioning the benefits of multitask learning.

Contribution

It introduces the application of ResNet-based models and pre-trained Conformers to vocal burst modeling, providing insights into multitask versus single-task training effectiveness.

Findings

ResNet models improved performance by 21.24% over baseline.

Pre-trained Conformer models achieved state-of-the-art results on paralinguistic tasks.

Single-task models outperformed multitask models in this setting.

Abstract

This technical report presents the modeling approaches used in our submission to the ICML Expressive Vocalizations Workshop & Competition multitask track (ExVo-MultiTask). We first applied image classification models of various sizes on mel-spectrogram representations of the vocal bursts, as is standard in sound event detection literature. Results from these models show an increase of 21.24% over the baseline system with respect to the harmonic mean of the task metrics, and comprise our team's main submission to the MultiTask track. We then sought to characterize the headroom in the MultiTask track by applying a large pre-trained Conformer model that previously achieved state-of-the-art results on paralinguistic tasks like speech emotion recognition and mask detection. We additionally investigated the relationship between the sub-tasks of emotional expression, country of origin, and age prediction, and discovered that the best performing models are trained as single-task models, questioning whether the problem truly benefits from a multitask setting.