TRACE: Temporal Relationship-Aware Conversational Entrainment Detection in Dyadic Speech

Source

Source name: arXiv
Source kind: paper
Authors / organization: Sathvik Manikantan Napa Ugandhar, Hao Zhang, Alison Gunzler, Yuzhe Wang, Thomas Thebaud, Georgi Tinchev, Venkatesh Ravichandran, Laureano Moro-Velázquez
Venue: arXiv: cs.CL
Published: 2026-06-29
Collected: 2026-06-30
URL: http://arxiv.org/abs/2606.30543v1
PDF: https://arxiv.org/pdf/2606.30543v1
DOI: UNKNOWN

Plain-Language Brief

Researchers introduce a dataset (DyadEE) for studying emotional entrainment in dyadic speech interactions, including both entrained and disrupted conversations. They also propose TRACE, a new framework that models interactions using acoustic embeddings from fine-tuned Whisper representations, which improves the accuracy of emotional entrainment detection compared to pooling utterances.

What the Document Says

Research is focused on emotional entrainment in conversational interactions between two people (dyadic speech).
DyadEE dataset is introduced for emotional entrainment detection, containing both entrained conversations and synthetic interactions with disrupted entrainment.
TRACE is a proposed window-level framework that models dyadic interaction as ordered sequences of acoustic embeddings derived from emotion fine-tuned Whisper representations.
Each sample in TRACE is treated as an 'interaction trace' rather than pooled utterances.
Experimental results on DyadEE show that incorporating conversational context and relationship information improves emotional entrainment detection.
TRACE achieves the best accuracy of 97.01% for emotional entrainment detection.

Key Concepts Mentioned

emotional entrainment
dyadic speech
DyadEE dataset
TRACE framework
acoustic embeddings
Whisper representations
accuracy

Claims Made by the Document

Improving emotional entrainment detection is important in speech AI agents.
Emotional entrainment is influenced by social relationships and conversational context.
DyadEE dataset improves the study of emotional entrainment in dyadic speech interactions.
TRACE improves emotional entrainment detection when compared to pooling utterances.
Incorporating conversational context and relationship information improves emotional entrainment detection.

Evidence Used

The text mentions the importance of understanding emotional entrainment in conversational interactions with AI agents.
The introduction of DyadEE dataset is explicitly stated for emotional entrainment detection.
TRACE is proposed as a means to improve emotional entrainment detection.
The text states that TRACE models interaction traces and improves accuracy over pooling utterances.
Experimental results on DyadEE show improvements when incorporating context and relationship information.

Methods / Approach

UNKNOWN

Limitations Stated by the Authors

UNKNOWN

Limitations Noted by the Agent

Agent interpretation — not from the source.

UNKNOWN

Open Questions

UNKNOWN

Notes for Future Q&A

May include interpretation. Labeled where it does.

UNKNOWN

Raw Abstract / Excerpt

With the proliferation of speech AI agents, understanding emotional entrainment in conversational interaction has become increasingly important. Emotional entrainment is shaped by social relationships and conversational context, influencing affective coordination over time. We introduce DyadEE, a dataset for emotional entrainment detection in dyadic speech interactions, containing both emotionally entrained conversations and synthetic interactions where entrainment is disrupted through partner swapping and emotion resynthesis. We further propose TRACE, a window-level framework that models dyadic interaction as ordered sequences of acoustic embeddings derived from emotion fine-tuned Whisper representations, treating each sample as an interaction trace rather than pooled utterances. Experimental results on DyadEE show that incorporating conversational context and relationship information improves emotional entrainment detection, with TRACE achieving the best accuracy of 97.01%.