Pith. sign in

REVIEW 3 major objections 4 minor 1 cited by

EmojiVoice: Towards long-term controllable expressivity in robot speech

T0 review · 3 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper claims that phrase-by-phrase emoji prompting lets a lightweight neural text-to-speech system keep a robot storyteller expressive over long interactions, while a fixed pleasant voice suits short assistant exchanges better.

desk verdict A genuinely useful open-source expressive TTS toolkit, but the headline storytelling result is confounded by fixed presentation order and needs a counterbalanced replication. read the letter →

arxiv 2506.15085 v2 pith:KAAEWOSP submitted 2025-06-18 cs.RO cs.HC

Paige Tuttösí , Shivam Mehta , Zachary Syvenky , Bermet Burkanova , Gustav Eje Henter , Angelica Lim This is my paper · ORCID
classification cs.ROcs.HC
keywords emojipromptingexpressivetext-to-speechrobotspeechtemporalvariabilitystorytellinghuman-robotinteractionMatcha-TTSsocialrobotics
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper proposes that varying vocal expressivity phrase by phrase, rather than using one fixed expressive voice, is what keeps a robot's speech engaging over long stretches. To make this practical, it introduces emoji prompting: a text-to-speech system fine-tuned to render each emoji as a distinct vocal style, with the emoji appended to each phrase selecting the style at synthesis time. The toolkit, built on the lightweight Matcha-TTS, runs in real time on robot-class hardware and can be customized with about three minutes of speech per style. The central finding is from a storytelling case study: listeners rated the emoji-varied voice as more expressive than a constant joyful voice, while the opposite held for a short assistant conversation.

What carries the argument

The central object is emoji prompting: appending one of eleven Unicode emojis to each text phrase and feeding that emoji token both to the text encoder and the flow-prediction network of Matcha-TTS, a non-autoregressive neural TTS trained with optimal-transport conditional flow matching. The fine-tuned checkpoint holds multiple voices in a single 78MB model, treating each emoji as a style label, so the system can switch expressive styles per phrase without reloading the model. This carries the argument by turning expressive control from a global voice setting into a per-phrase local control that costs almost nothing at inference and works naturally with LLM-generated text that already contains emojis.

What would settle it

Run the same storytelling script with the emoji-varied voice but with emojis randomly reassigned to phrases; if perceived expressiveness drops to the level of the constant pleasant voice, the effect depends on emoji-content matching rather than temporal variation. Alternatively, measure acoustic prosodic features such as pitch range and speaking rate across the 11 emoji conditions; if they do not separate into distinct clusters, the model has not actually learned 11 distinct styles.

Watch

Extended reading notes

Core claim

On its own terms, the paper's central claim is that phrase-by-phrase emoji prompting of a fine-tuned Matcha-TTS creates long-term temporally variable expressive speech that listeners perceive as more expressive than a single joyful expressive voice in storytelling. In the storytelling case study, the emoji-varied voice scored significantly higher than the baseline on expressiveness, social impression, and suitability, and significantly higher than the pleasant voice on expressiveness, and it was the first-choice voice with high confidence. In the assistant conversation, the pleasant voice was preferred and rated more suitable, showing that the appropriate vocal strategy depends on the task. The claimed mechanism is that variation itself prevents the monotony of a constant expressive style over long text, even though a consistently pleasant voice can be more appropriate for short, task-oriented utterances.

Load-bearing premise

The load-bearing premise is that fine-tuning Matcha-TTS on roughly 40 sentences per emoji (about three minutes of speech) per speaker yields 11 distinct, consistent, and controllable vocal styles, so the reported perceptual differences come from those learned styles rather than from the specific emoji choices, voice quality, or presentation order.

Editorial extensions

If this is right

  • Social robots can vary vocal expression per phrase on-board, without a cloud GPU, using a model with about 20.9M parameters and a real-time factor below one.
  • Long storytelling interactions can be made more engaging with an emoji-varied voice than with a fixed joyful voice, so robot voice design should treat expression as dynamic rather than a single setting.
  • For assistant tasks, a consistently pleasant voice may be more appropriate than high variability, so the task and interaction context should determine which expressive strategy to deploy.
  • LLM-based conversational agents can append emojis to route voice styles automatically, making expressive control nearly free for autonomous spoken dialogue systems.
  • The toolkit lowers the data cost of custom expressive voices to roughly three minutes of speech per style, which could let roboticists build task-specific expressive voices quickly.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the benefit comes from temporal variation rather than from the specific emoji set, then randomly varying a small set of pleasant styles might reproduce part of the storytelling gain; this is a testable extension the paper does not run.
  • The one-emoji-per-phrase granularity is a ceiling on control; word-level emoji markup could allow emphasis on individual words and might address the participant report that timing, pitch, and emphasis sometimes felt off.
  • Because the LLM's emoji choices strongly shaped perceived appropriateness in the autonomous case, future systems should couple emoji selection to conversational content rather than treating emojis as purely style-only tokens.
  • The fixed presentation order in the first two case studies means the reported preference differences could partly reflect ordering or fatigue effects; a fully randomized within-subject design would test whether the effect persists.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The paper presents EmojiVoice, an open-source text-to-speech toolkit built on Matcha-TTS, and proposes emoji prompting as a method for phrase-level control of expressivity in robot speech. It reports three case studies: a scripted robot assistant conversation (Case Study 1), a storytelling task (Case Study 2), and an autonomous speech-to-speech interactive agent (Case Study 3). The central claim is that phrase-by-phrase emoji variation improves perceived expressivity relative to a constant 'joyful' voice in storytelling, while varied expressivity is not preferred in the assistant use case. The paper also documents the fine-tuning recipe, reports real-time performance, and releases code, checkpoints, and data.

Significance. If the empirical claim were established, the paper would offer a practical and timely contribution to HRI: a lightweight, open-source, controllable TTS system that runs in real time on modest hardware. The release of checkpoints, training scripts, and documentation is a concrete strength, as is the attempt to evaluate the system in multiple robot interaction contexts. However, the central empirical claim is currently undermined by a fixed voice order in Case Study 2 and by reuse of the same participant pool across Case Studies 1 and 2. The toolkit contribution remains plausible, but the validation of the headline storytelling result needs substantial improvement.

major comments (3)
  1. [IV-B.1 and IV-B.2] The storytelling result that carries the paper's central claim is confounded by presentation order. Section IV-B.1 states 'The order of the voices were: Pleasant, Baseline, Emoji,' so the Emoji condition was always the last of the three within-subject presentations; Section IV-B.2 reports that Case Study 2 'used the same participant pool as Case Study 1,' so all 24 participants had already heard and rated the same voices in a different task. The observed pattern across the two studies is exactly what an order or contrast effect would predict: in Case Study 1, with order Baseline, Emoji, Pleasant, the Pleasant voice was preferred; in Case Study 2, with order Pleasant, Baseline, Emoji, the Emoji voice was preferred. A counterbalanced design, a fresh participant pool, or a direct statistical test for order effects is required before the perceived-expressivity advantage can be attributed to emoji prompting.
  2. [IV-A.3 and IV-B.3] The statistical analyses in Tables II and III appear to treat the data as independent observations even though every participant rated all three voices within each case study, and the same participants were reused across Case Studies 1 and 2. If a between-subjects one-way ANOVA was used, it violates the independence assumption; if a repeated-measures ANOVA was intended, the paper should report the full model, including error degrees of freedom, sphericity corrections, and participant as a random effect. As reported, the p-values and omega-squared values for xMOS may be overstated. A mixed-effects model or a repeated-measures analysis is needed to support the reported differences.
  3. [III-C] The paper's interpretation of every comparative result assumes that the 11 emoji voices are distinct, internally consistent, and different from the baseline and Pleasant voices, yet no acoustic verification or perceptual discrimination test is reported. The fine-tuning uses only about 40 sentences per emoji per speaker, so it is possible that some emoji conditions simply captured recording variation or training noise. The claim that emoji prompting produces distinct and controllable vocal styles would be substantially strengthened by an objective analysis of the synthesized voices, such as F0, duration, spectral measures, or a style-classification test.
minor comments (4)
  1. [Abstract] The abstract says 'fine-grained control of expressivity on a phase level'; this should be 'phrase level' to match the terminology used in the rest of the paper.
  2. [IV-A.1 and IV-B.1] Case Study 1 also uses a fixed voice order (Baseline, Emoji, Pleasant), so the conclusion that Pleasant is preferred in the assistant task is subject to the same order-confound concern; this should at least be acknowledged in the limitations.
  3. [III-C and IV-B.2] The paper says in Section III-C that researchers can use 'as little as 3 minutes of speech per emoji' but the data collection section describes 40 training sentences per emoji; the relationship between these two numbers should be clarified.
  4. [Tables II and III] The Tukey HSD notation in the tables is dense; expanding the caption with a worked example of the 'E-P' notation, or providing a separate pairwise-comparison table, would improve readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: the perceptual claims are evaluated by a held-out user study, and the only self-citation (Matcha-TTS) is an external, code-reproduced component.

full rationale

EmojiVoice makes no formal derivation; its central claim is empirical. The abstract's assertion that 'varied emoji prompting improved the perception and expressivity of speech over a long period in a storytelling task' is tested by user ratings of held-out story scripts, not by construction from training data. The 'Emoji' condition is a TTS configuration, not a renamed version of the rating outcome, and the experiment could have failed: in Case Study 1 the 'Pleasant' voice was preferred, so the perceptual result is not forced. The only self-citation is Matcha-TTS, which is an external open-source codebase used as a backbone rather than as evidence for the perceptual claim; under the stated rules, a code-reproduced component counts as independent support. The use of the same emoji set in training and evaluation is a manipulation-check concern, not a circular reduction, and the fixed presentation order is a methodological confound rather than a definitional equivalence. No step reduces a prediction to its own input, so the paper is not circular.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The paper's claims rest on empirical assumptions about emoji semantics, fine-tuning efficacy with small data, and subjective rating validity. These are standard domain assumptions for HRI/TTS work, but they are not independently verified beyond the reported user studies.

assumptions (4)
  • domain assumption Emojis are a sufficiently rich and reliable representation of emotional and social expressivity for robot speech.
    The paper relies on emojis as both a data-collection prompt and a runtime control signal; Section III-B motivates this with the Unicode emoji catalog and LLM familiarity, but does not validate that listeners associate the 11 chosen emojis with the intended vocal styles.
  • domain assumption Fine-tuning a multi-speaker TTS on a small dataset (50 phrases per emoji, 3 speakers) yields distinct, stable voice styles.
    Section III-C describes 20-minute fine-tuning runs and claims 'as little as 3 minutes of speech per emoji' suffices, but no objective acoustic or classifier-based verification of style distinctness is provided.
  • domain assumption Subjective MOS-type ratings on 10-point Likert scales are valid measures of expressivity, social impression, and suitability over short interactions.
    Section IV-A.1 uses MOS-X2 and xMOS scales; the paper does not calibrate these ratings against any external anchor or include inter-rater reliability.
  • domain assumption The robotic embodiment and accompanying animations do not differentially bias the voice ratings across conditions.
    The same animations were used for all voices within each robot, but the paper does not control for voice-animation matching, which could favor the pleasant or emoji voices.

how reviews work

0 comments
Cite this review

Pith. "Pith review of EmojiVoice: Towards long-term controllable expressivity in robot speech." pith.science (2026). https://pith.science/paper/KAAEWOSP

@misc{pith2026250615085,
  author       = {Pith},
  title        = {Pith review of: EmojiVoice: Towards long-term controllable expressivity in robot speech},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KAAEWOSP}},
  note         = {Machine review of arXiv:2506.15085}
}
read the original abstract

Humans vary their expressivity when speaking for extended periods to maintain engagement with their listener. Although social robots tend to be deployed with ``expressive'' joyful voices, they lack this long-term variation found in human speech. Foundation model text-to-speech systems are beginning to mimic the expressivity in human speech, but they are difficult to deploy offline on robots. We present EmojiVoice, a free, customizable text-to-speech (TTS) toolkit that allows social roboticists to build temporally variable, expressive speech on social robots. We introduce emoji-prompting to allow fine-grained control of expressivity on a phase level and use the lightweight Matcha-TTS backbone to generate speech in real-time. We explore three case studies: (1) a scripted conversation with a robot assistant, (2) a storytelling robot, and (3) an autonomous speech-to-speech interactive agent. We found that using varied emoji prompting improved the perception and expressivity of speech over a long period in a storytelling task, but expressive voice was not preferred in the assistant use case.

Figures

Figures reproduced from arXiv: 2506.15085 by the authors.

Figure 1
Figure 1. Miroka robot speaking with EmojiVoice expressive TTS. [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Matcha-TTS architecture [11] with the addition of emoji prompting [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 4
Figure 4. Example of text with emojis. Sample of the script provided to the [PITH_FULL_IMAGE:figures/full_fig_p003_4.png] view at source ↗
Figures from the paper (4 more)
Figure 8
Figure 8. Figure 8: Case Study 1: Robot Helper with human “Alex”, script excerpt. [PITH_FULL_IMAGE:figures/full_fig_p004_8.png]
Figure 10
Figure 10. Figure 10: Case Study 1: Robot Helper. Counts for voice preference [PITH_FULL_IMAGE:figures/full_fig_p005_10.png]
Figure 11
Figure 11. Figure 11: Case Study 2: Storytelling. Counts for voice preference [PITH_FULL_IMAGE:figures/full_fig_p006_11.png]
Figure 12
Figure 12. Figure 12: Case Study 3: Story Building, example interaction. [PITH_FULL_IMAGE:figures/full_fig_p007_12.png]

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. The Prosody of Emojis

    cs.CL 2025-08 conditional novelty 6.0 of 10

    Speakers systematically modify prosody when reading sentences with different emojis, and listeners can recover the intended emoji from prosody alone, with larger semantic differences yielding larger prosodic shifts.

Reference graph

Works this paper leans on

48 extracted references · 46 canonical work pages · cited by 1 Pith paper

  1. [1]

    Expresso: transformation of expressivity in speech,

    G. Beller, “Expresso: transformation of expressivity in speech,” in Speech Prosody, 2010, p. paper 043

    work page 2010

  2. [2]

    What makes a good conversation? challenges in designing truly conversational agents,

    L. Clarket al., “What makes a good conversation? challenges in designing truly conversational agents,” inCHI, 2019, pp. 1–12

    work page 2019

  3. [3]

    Describing the emotional states that are expressed in speech,

    R. Cowie and R. R. Cornelius, “Describing the emotional states that are expressed in speech,”Speech communication, vol. 40, no. 1-2, pp. 5–32, 2003

    work page 2003

  4. [4]

    Intonation in robot speech: Does it work the same as with people?

    E. Velner, P. P. Boersma, and M. M. de Graaf, “Intonation in robot speech: Does it work the same as with people?” inHRI, 2020, p. 569–578

    work page 2020

  5. [5]

    Flat vs. expressive storytelling: Young children’s learning and retention of a social robot’s narrative,

    K. Westlundet al., “Flat vs. expressive storytelling: Young children’s learning and retention of a social robot’s narrative,”Frontiers in human neuroscience, vol. 11, p. 295, 2017

    work page 2017

  6. [6]

    Nonverbal immediacy as a characterisation of social behaviour for human–robot interaction,

    J. Kennedy, P. Baxter, and T. Belpaeme, “Nonverbal immediacy as a characterisation of social behaviour for human–robot interaction,” International Journal of Social Robotics, vol. 9, pp. 109–128, 2017

    work page 2017

  7. [7]

    The impact and implications of virtual character expressiveness on learning and agent–learner interactions,

    G. Veletsianos, “The impact and implications of virtual character expressiveness on learning and agent–learner interactions,”Journal of Computer Assisted Learning, vol. 25, no. 4, pp. 345–357, 2009

    work page 2009

  8. [8]

    The effect of an artificial agent’s vocal expressiveness on immediacy and learning,

    S. Fountoukidou, U. Matzat, J. Ham, and C. Midden, “The effect of an artificial agent’s vocal expressiveness on immediacy and learning,” Journal of Computer Assisted Learning, vol. 38, no. 2, pp. 500–512, 2022

    work page 2022

Show all 48 references
  1. [9]

    Expressive synthetic voices: Consider- ations for human robot interaction,

    S. Hennig and R. Chellali, “Expressive synthetic voices: Consider- ations for human robot interaction,” in2012 IEEE RO-MAN: The 21st IEEE International Symposium on Robot and Human Interactive Communication. IEEE, 2012, pp. 589–595

    2012

  2. [10]

    What makes a charismatic speaker? a computer-based acoustic-prosodic analysis of steve jobs tone of voice,

    O. Niebuhr, J. V oße, and A. Brem, “What makes a charismatic speaker? a computer-based acoustic-prosodic analysis of steve jobs tone of voice,”Computers in Human Behavior, vol. 64, pp. 366–382, 2016

    2016

  3. [11]

    Matcha- TTS: A fast TTS architecture with conditional flow matching,

    S. Mehta, R. Tu, J. Beskow, ´E. Sz ´ekely, and G. E. Henter, “Matcha- TTS: A fast TTS architecture with conditional flow matching,” in ICASSP, 2024

    2024

  4. [12]

    Emosphere-tts: Emotional style and intensity modeling via spherical emotion vector for controllable emotional text-to-speech,

    D.-H. Cho, H.-S. Oh, S.-B. Kim, S.-H. Lee, and S.-W. Lee, “Emosphere-tts: Emotional style and intensity modeling via spherical emotion vector for controllable emotional text-to-speech,” inInter- speech 2024, 2024, pp. 1810–1814

    2024

  5. [13]

    Controlling emotion in text-to-speech with natural language prompts,

    T. Bott, F. Lux, and N. T. Vu, “Controlling emotion in text-to-speech with natural language prompts,” inInterspeech, 2024, pp. 1795–1799

    2024

  6. [14]

    Generating expressive speech for storytelling applications,

    M. Theune, K. Meijs, D. Heylen, and R. Ordelman, “Generating expressive speech for storytelling applications,”Trans. Audio Speech Lang. Process., vol. 14, no. 4, pp. 1137–1144, 2006

    2006

  7. [15]

    Fun or boring? a web-based evaluation of expressive synthesis for children,

    K. Gustafson and D. House, “Fun or boring? a web-based evaluation of expressive synthesis for children,” inEurosepeech, 2001, pp. 565– 568

    2001

  8. [16]

    Contextspeech: Expressive and efficient text-to-speech for paragraph reading,

    Y . Xiaoet al., “Contextspeech: Expressive and efficient text-to-speech for paragraph reading,” inINTERSPEECH 2023, 2023, pp. 4883–4887

    2023

  9. [17]

    Storytts: A highly expressive text-to-speech dataset with rich textual expressiveness annotations,

    S. Liu, Y . Guo, X. Chen, and K. Yu, “Storytts: A highly expressive text-to-speech dataset with rich textual expressiveness annotations,” in ICASSP, 2024, pp. 11 521–11 525

    2024

  10. [18]

    Reliability and security of ai hardware,

    D. Gnad and et al., “Reliability and security of ai hardware,” inETS, 2024, pp. 1–10

    2024

  11. [19]

    Artificial intelligence for long-term robot autonomy: A survey,

    L. Kunze, N. Hawes, T. Duckett, M. Hanheide, and T. Krajn ´ık, “Artificial intelligence for long-term robot autonomy: A survey,”RA-L, vol. 3, no. 4, pp. 4023–4030, 2018

    2018

  12. [20]

    Pervasive ai for iot applications: A survey on resource-efficient distributed artificial intelligence,

    E. Baccouret al., “Pervasive ai for iot applications: A survey on resource-efficient distributed artificial intelligence,”IEEE Communi- cations Surveys and Tutorials, vol. 24, no. 4, pp. 2366–2418, 2022

    2022

  13. [21]

    What’s up with privacy? user preferences and privacy concerns in intelligent personal assistants,

    L. Manikonda, A. Deotale, and S. Kambhampati, “What’s up with privacy? user preferences and privacy concerns in intelligent personal assistants,” inAIES, 2018, p. 229–235

    2018

  14. [22]

    Survey on threats and risks in the cloud computing environment,

    M. Abdurachman, E. Abdurachman, F. L. Gaol, and B. Soewito, “Survey on threats and risks in the cloud computing environment,” Procedia Computer Science, vol. 161, pp. 1325–1332, 2019

    2019

  15. [23]

    V oicecraft: Zero-shot speech editing and text-to-speech in the wild,

    P. Peng, P.-Y . Huang, A. Mohamed, and D. Harwath, “V oicecraft: Zero-shot speech editing and text-to-speech in the wild,”arXiv, 2024

    2024

  16. [24]

    Coquixtts,

    Coqui-AI, “Coquixtts,”https://github.com/coqui-ai/TTS/tree/dev, 2021

    2021

  17. [25]

    Natural language guidance of high-fidelity text- to-speech with synthetic annotations,

    D. Lyth and S. King, “Natural language guidance of high-fidelity text- to-speech with synthetic annotations,”arXiv, 2024

    2024

  18. [26]

    Fastpitch: Parallel text-to-speech with pitch prediction,

    A. Ła ´ncucki, “Fastpitch: Parallel text-to-speech with pitch prediction,” inICASSP, 2021, pp. 6588–6592

    2021

  19. [27]

    Scaling up online speech recognition using convnets,

    V . Pratap and et al., “Scaling up online speech recognition using convnets,” inInterspeech 2020, 2020, pp. 3376–3380

    2020

  20. [28]

    A systematic review of emoji: Current research and future perspectives,

    Q. Bai, Q. Dan, Z. Mu, and M. Yang, “A systematic review of emoji: Current research and future perspectives,”Frontiers in Psychology, vol. 10, 2019

    2019

  21. [29]

    First suggestions for an emotion annotation and representation language,

    M. Schr ¨oder, H. Pirker, and M. Lamolle, “First suggestions for an emotion annotation and representation language,” inLREC, vol. 6, 2006, pp. 88–92

    2006

  22. [30]

    Emotion knowledge: Further exploration of a prototype approach,

    P. Shaver, J. Schwartz, D. Kirson, and C. O’Connor, “Emotion knowledge: Further exploration of a prototype approach,”Journal of personality and social psychology, vol. 52, no. 6, pp. 1061–1086, 1987

    1987

  23. [31]

    Chapter 1 - a general psychoevolutionary theory of emotion,

    R. Plutchik, “Chapter 1 - a general psychoevolutionary theory of emotion,” inTheories of Emotion, R. Plutchik and H. Kellerman, Eds. Academic Press, 1980, pp. 3–33

    1980

  24. [32]

    Emojineering part 1: Machine learning for emoji trends,

    T. Dimson, “Emojineering part 1: Machine learning for emoji trends,” Instagram Engineering Blog, vol. 30, 2015

    2015

  25. [33]

    Real-time message sentiment augmentation by emoji symbols,

    G. Savage and A. Yung, “Real-time message sentiment augmentation by emoji symbols,” 2024

    2024

  26. [34]

    How good is gpt’s “emojinal intel- ligence

    M. Dunn and K. Hopkinson, “How good is gpt’s “emojinal intel- ligence”? investigating emoji patterns in llm-generated social media text,” inProceedings of the International Conference on AI Research. Academic Conferences and publishing limited, 2024

    2024

  27. [35]

    Grounding conversational robots on vision through dense captioning and large language models,

    L. Grassi, Z. Hong, C. T. Recchiuto, and A. Sgorbissa, “Grounding conversational robots on vision through dense captioning and large language models,” inICRA, 2024, pp. 5492–5498

    2024

  28. [36]

    Integrating visual context into language models for situated social conversation starters,

    R. Janssens, P. Wolfert, T. Demeester, and T. Belpaeme, “Integrating visual context into language models for situated social conversation starters,”IEEE Transactions on Affective Computing, pp. 1–14, 2024

    2024

  29. [37]

    Large language models for human–robot interaction: A review,

    C. Zhang, J. Chen, J. Li, Y . Peng, and Z. Mao, “Large language models for human–robot interaction: A review,”Biomimetic Intelligence and Robotics, vol. 3, no. 4, p. 100131, 2023

    2023

  30. [38]

    Language models for human- robot interaction,

    E. Billing, J. Ros ´en, and M. Lamb, “Language models for human- robot interaction,” inHRI, ser. HRI ’23. Association for Computing Machinery, 2023, p. 905–906

    2023

  31. [39]

    Robust speech recognition via large-scale weak super- vision,

    A. Radford, J. W. Kim, T. Xu, G. Brockman, C. McLeavey, and I. Sutskever, “Robust speech recognition via large-scale weak super- vision,” inICML, 2023

    2023

  32. [40]

    Gaze and filled pause detection for smooth human-robot conversations,

    M. Bilac, M. Chamoux, and A. Lim, “Gaze and filled pause detection for smooth human-robot conversations,” inHumanoids, 2017, pp. 297– 304

    2017

  33. [41]

    The llama 3 herd of models,

    A. Dubey and et al., “The llama 3 herd of models,” 2024

    2024

  34. [42]

    Investigating mos-x ratings of synthetic and human voices,

    J. R. Lewis and IBM.Corp, “Investigating mos-x ratings of synthetic and human voices,”Voice Interaction Design, vol. 2, no. 1, p. 22, 2018

    2018

  35. [43]

    Word-level text markup for prosody control in speech synthesis,

    Y . Korotkova, I. Kalinovskiy, and T. Vakhrusheva, “Word-level text markup for prosody control in speech synthesis,” inInterspeech, 2024, pp. 2280–2284

    2024

  36. [44]

    Giving robots a voice: Human-in-the-loop voice creation and open- ended labeling,

    P. van Rijn, S. Mertes, K. Janowski, K. Weitz, N. Jacoby, and E. Andr´e, “Giving robots a voice: Human-in-the-loop voice creation and open- ended labeling,” inCHI, 2024

    2024

  37. [45]

    Spoken language interaction with robots: Recom- mendations for future research,

    M. Margeet al., “Spoken language interaction with robots: Recom- mendations for future research,”Computer Speech and Language, vol. 71, p. 101255, 2022

    2022

  38. [46]

    Improving the quality of neural tts using long-form content and multi-speaker multi-style modeling,

    T. Raitio, J. Latorre, A. Davis, T. Morrill, and L. Golipour, “Improving the quality of neural tts using long-form content and multi-speaker multi-style modeling,” inSSW, 2023

    2023

  39. [47]

    No more mumbles: Enhancing robot intelligibility through speech adaptation,

    Q. Ren, Y . Hou, D. Botteldooren, and T. Belpaeme, “No more mumbles: Enhancing robot intelligibility through speech adaptation,” RA-L, vol. 9, no. 7, pp. 6162–6169, 2024

    2024

  40. [48]

    Mmm whatcha say? uncovering distal and proximal context effects in first and second-language word perception using psychophysical reverse correlation,

    P. Tutt ¨os´ıet al., “Mmm whatcha say? uncovering distal and proximal context effects in first and second-language word perception using psychophysical reverse correlation,” inInterspeech, 2024, pp. 1010– 1014

    2024

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.