Dual Information Speech Language Models for Emotional Conversations

TL;DR

This paper introduces a novel speech-language model with heterogeneous adapters and a weakly supervised training strategy to better interpret emotional speech by disentangling paralinguistic and linguistic cues, improving contextual understanding.

Contribution

It proposes a new model architecture and training method that effectively separates paralinguistic and linguistic information in speech-language models.

Findings

Achieves competitive performance in emotional conversation tasks.

Effectively disentangles paralinguistic and linguistic cues.

Maintains contextual understanding while interpreting speech.

Abstract

Conversational systems relying on text-based large language models (LLMs) often overlook paralinguistic cues, essential for understanding emotions and intentions. Speech-language models (SLMs), which use speech as input, are emerging as a promising solution. However, SLMs built by extending frozen LLMs struggle to capture paralinguistic information and exhibit reduced context understanding. We identify entangled information and improper training strategies as key issues. To address these issues, we propose two heterogeneous adapters and suggest a weakly supervised training strategy. Our approach disentangles paralinguistic and linguistic information, enabling SLMs to interpret speech through structured representations. It also preserves contextual understanding by avoiding the generation of task-specific vectors through controlled randomness. This approach trains only the adapters on common datasets, ensuring parameter and data efficiency. Experiments demonstrate competitive performance in emotional conversation tasks, showcasing the model's ability to effectively integrate both paralinguistic and linguistic information within contextual settings.