<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="liu-etal-2026-emotra">
    <titleInfo>
        <title>EmoTra-TTS: Smooth Intra-Utterance Emotion Transitions for Speech Synthesis</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Tianchi</namePart>
        <namePart type="family">Liu</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Zeyang</namePart>
        <namePart type="family">Song</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Tianrui</namePart>
        <namePart type="family">Wang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Zhipeng</namePart>
        <namePart type="family">Li</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Chenglin</namePart>
        <namePart type="family">Xu</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Steven</namePart>
        <namePart type="given">Y</namePart>
        <namePart type="family">Guo</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2026-10</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Hung-Yi</namePart>
            <namePart type="family">Lee</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Sunipa</namePart>
            <namePart type="family">Dev</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Jessy</namePart>
            <namePart type="family">Li</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Desmond</namePart>
            <namePart type="family">Elliott</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Budapest, Hungary</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-459-0</identifier>
    </relatedItem>
    <abstract>Psychological research on emotion dynamics has established that human affect is a continuous, evolving process: emotions rise, decay, and transition within seconds. Current emotional text-to-speech (TTS) systems, however, condition on a single discrete label or static embedding per utterance, fundamentally misaligning with the temporal nature of affect. While recent LLM-based TTS systems may implicitly vary prosody through text understanding, such variation is neither explicitly controllable nor precise enough for targeted intra-utterance transitions. We address three challenges: (1) a multi-pass flow blending pipeline synthesizes frame-aligned transition audio, circumventing the scarcity of natural intra-utterance transitions; (2) dual-stage Valence–Arousal–Dominance (VAD) conditioning guides prosodic planning in the LLM and acoustic realization in the flow decoder via frame-level VAD embeddings; (3) direction–magnitude decoupled injection structurally separates emotion direction from injection magnitude, preventing content degradation. EmoTra-TTS adds only +0.43% parameters with no latency overhead, achieves 30%–87% relative improvement on emotion transition quality, corroborated by 67.5%–79.5% overall win rates in a pairwise preference test against four SOTA baselines. The complete code, including data synthesis and training pipelines, will be released after the review process.</abstract>
    <identifier type="citekey">liu-etal-2026-emotra</identifier>
    <location>
        <url>https://aclanthology.org/2026.emnlp-main.736/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>17073</start>
            <end>17091</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T EmoTra-TTS: Smooth Intra-Utterance Emotion Transitions for Speech Synthesis
%A Liu, Tianchi
%A Song, Zeyang
%A Wang, Tianrui
%A Li, Zhipeng
%A Xu, Chenglin
%A Guo, Steven Y.
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F liu-etal-2026-emotra
%X Psychological research on emotion dynamics has established that human affect is a continuous, evolving process: emotions rise, decay, and transition within seconds. Current emotional text-to-speech (TTS) systems, however, condition on a single discrete label or static embedding per utterance, fundamentally misaligning with the temporal nature of affect. While recent LLM-based TTS systems may implicitly vary prosody through text understanding, such variation is neither explicitly controllable nor precise enough for targeted intra-utterance transitions. We address three challenges: (1) a multi-pass flow blending pipeline synthesizes frame-aligned transition audio, circumventing the scarcity of natural intra-utterance transitions; (2) dual-stage Valence–Arousal–Dominance (VAD) conditioning guides prosodic planning in the LLM and acoustic realization in the flow decoder via frame-level VAD embeddings; (3) direction–magnitude decoupled injection structurally separates emotion direction from injection magnitude, preventing content degradation. EmoTra-TTS adds only +0.43% parameters with no latency overhead, achieves 30%–87% relative improvement on emotion transition quality, corroborated by 67.5%–79.5% overall win rates in a pairwise preference test against four SOTA baselines. The complete code, including data synthesis and training pipelines, will be released after the review process.
%U https://aclanthology.org/2026.emnlp-main.736/
%P 17073-17091

Markdown (Informal)

[EmoTra-TTS: Smooth Intra-Utterance Emotion Transitions for Speech Synthesis](https://aclanthology.org/2026.emnlp-main.736/) (Liu et al., EMNLP 2026)

• EmoTra-TTS: Smooth Intra-Utterance Emotion Transitions for Speech Synthesis (Liu et al., EMNLP 2026)

ACL

• Tianchi Liu, Zeyang Song, Tianrui Wang, Zhipeng Li, Chenglin Xu, and Steven Y. Guo. 2026. EmoTra-TTS: Smooth Intra-Utterance Emotion Transitions for Speech Synthesis. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 17073–17091, Budapest, Hungary. Association for Computational Linguistics.