MATER: Multi-level Acoustic and Textual Emotion Representation for Interpretable Speech Emotion Recognition

TL;DR

This paper introduces MATER, a hierarchical framework combining acoustic and textual features at multiple levels for more interpretable speech emotion recognition in naturalistic conditions, addressing variability and ambiguity.

Contribution

The paper proposes a novel multi-level hierarchical approach integrating acoustic and textual cues, along with an uncertainty-aware ensemble, to improve speech emotion recognition robustness.

Findings

MATER achieved 41.01% Macro-F1 in emotion recognition.

Secured second place with a CCC of 0.6941 in valence prediction.

Demonstrated effectiveness in handling natural speech variability.

Abstract

This paper presents our contributions to the Speech Emotion Recognition in Naturalistic Conditions (SERNC) Challenge, where we address categorical emotion recognition and emotional attribute prediction. To handle the complexities of natural speech, including intra- and inter-subject variability, we propose Multi-level Acoustic-Textual Emotion Representation (MATER), a novel hierarchical framework that integrates acoustic and textual features at the word, utterance, and embedding levels. By fusing low-level lexical and acoustic cues with high-level contextualized representations, MATER effectively captures both fine-grained prosodic variations and semantic nuances. Additionally, we introduce an uncertainty-aware ensemble strategy to mitigate annotator inconsistencies, improving robustness in ambiguous emotional expressions. MATER ranks fourth in both tasks with a Macro-F1 of 41.01% and an average CCC of 0.5928, securing second place in valence prediction with an impressive CCC of 0.6941.