pith. sign in

arxiv: 2605.20200 · v1 · pith:6RSFY3DXnew · submitted 2026-04-06 · 💻 cs.HC · cs.AI

Evaluating multimodal emotion recognition in proactive conversational agents: A user study

Adnana Dragut , Raquel Lacuesta , F. Xavier Gaya-Morey , Jose M. Buades-Rubio This is my paper

Pith reviewed 2026-05-21 09:49 UTC · model grok-4.3

classification 💻 cs.HC cs.AI
keywords multimodal emotion recognitionproactive conversational agentssocially interactive agentspoker face effectlinguistic analysisfacial recognitionuser studygenerative AI
0
0 comments X

The pith

In conversations with proactive AI agents, users maintain serious facial expressions even during positive emotions, making generative linguistic analysis more reliable than visual cues.

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

This paper evaluates a multimodal emotion recognition system built into a generative-AI-powered proactive socially interactive agent. The system combines real-time facial recognition via computer vision with semantic analysis of user language. A study of twenty participants in unscripted dialogues found that visual cues frequently mismatched internal states because users adopted a concentrated, serious expression regardless of positive feelings. Linguistic analysis performed better by interpreting verbal context. The work also shows that the agent can deliberately elicit emotions through theme changes and empathetic or humorous phrasing, yet overly abrupt proactivity sometimes produced disengagement and a sense of artificiality.

Core claim

When users interacted with the proactive agent, they consistently displayed serious and concentrated facial expressions even while reporting positive internal emotions. This produced a clear discrepancy between automated visual readings and actual affective states. As a result, the generative-AI linguistic analysis engine proved significantly more reliable because it could place verbal expressions in conversational context. The study further demonstrated that the agent could elicit targeted emotions by shifting conversational themes and employing structured linguistic patterns such as empathy or humor, although uncalibrated proactivity occasionally led to user disengagement and perceptions 0

What carries the argument

Multimodal emotion recognition module that fuses computer-vision facial analysis with generative-AI semantic linguistic analysis.

If this is right

  • Proactive agents can reliably elicit specific emotions by adapting conversational themes and using empathetic or humorous language patterns.
  • Overly abrupt proactivity risks user disengagement and perceptions of artificiality.
  • Future SIAs should prioritize deep linguistic context when tracking emotional states to support more natural interactions.
  • Dynamic adaptation to the user's evolving affective state improves the quality of human-like dialogue.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Visual emotion recognition may be systematically less effective in human-AI settings than in human-human settings due to self-presentation effects.
  • Hybrid systems could improve by weighting linguistic signals more heavily whenever the user is known to be interacting with an artificial agent.
  • Longer-term deployments might reveal whether users habituate to the poker-face pattern or whether it persists across repeated sessions.
  • The same linguistic-priority approach could be tested in other proactive AI domains such as tutoring or mental-health support agents.

Load-bearing premise

Internal emotional states can be measured independently and accurately enough to establish a reliable discrepancy with the visual cues observed.

What would settle it

A replication study in which participants' self-reported emotions during AI dialogues match their detected facial expressions at rates comparable to human-human conversations.

read the original abstract

This article presents a multimodal emotion recognition module integrated into a proactive Socially Interactive Agent (SIA) powered by generative artificial intelligence. The system evaluates real-time affective states through two distinct channels: a computer vision-based facial recognition module and a semantic linguistic analysis engine. To validate the framework, an empirical study was conducted with 20 users who engaged in dynamic, unscripted dialogues with the conversational agent. The findings reveal a significant discrepancy between automated visual cues and actual internal emotional states. When interacting with the AI, users consistently exhibited a "poker face" effect, displaying serious, concentrated facial expressions even when experiencing positive emotions. Consequently, the generative AI linguistic analysis proved significantly more reliable, by contextualizing the users' verbal expressions. Furthermore, an analysis of the interaction dynamics demonstrated that SIAs can effectively elicit specific emotions by adapting conversational themes and employing structured linguistic patterns, such as empathetic or humorous language. However, the study also noted that instances of uncalibrated proactivity occasionally led to user disengagement and a perception of artificiality. Ultimately, this research highlights the necessity of refining SIAs to dynamically adapt to users' emotional evolution, relying on deep linguistic context to foster more natural, human-like interactions.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript presents a multimodal emotion recognition module for a proactive Socially Interactive Agent (SIA) powered by generative AI, combining computer vision-based facial recognition with semantic linguistic analysis. A user study with 20 participants in unscripted dialogues reports a 'poker face' effect in which users display serious expressions despite positive internal emotions, rendering linguistic analysis more reliable than visual cues. The work also examines how SIAs elicit specific emotions via adaptive themes and linguistic patterns, while noting that uncalibrated proactivity can cause disengagement and perceptions of artificiality.

Significance. If the empirical results hold after methodological clarification, the paper contributes to HCI by demonstrating limitations of visual emotion recognition during task-focused interactions and the comparative strength of linguistic context for affective inference. This has implications for designing emotionally adaptive conversational agents. The use of real-user unscripted dialogues is a positive aspect, though the small sample and absence of detailed validation protocols limit the strength of the claims.

major comments (2)
  1. Abstract and study description: The central claim of a significant discrepancy between visual cues and 'actual internal emotional states' (including the poker face effect and superior reliability of linguistic analysis) depends on an independent ground-truth measure of participants' emotions. No protocol is described for obtaining this measure (e.g., self-report timing, validation against physiological markers, or controls for demand characteristics and social-desirability bias), which is load-bearing for the reliability comparison.
  2. User study section: The findings on emotion elicitation and disengagement are presented without reported statistical tests, effect sizes, error bars, baseline comparisons, or justification for the sample size of 20, making it difficult to evaluate whether the observed patterns support the stated conclusions about modality reliability and proactivity effects.
minor comments (2)
  1. Abstract: Consider specifying the exact conversational themes or structured linguistic patterns (empathetic/humorous) used to elicit emotions, as this would strengthen the reproducibility of the interaction dynamics findings.
  2. Overall: The term 'uncalibrated proactivity' is used without a clear operational definition; adding one would improve clarity for readers.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive and detailed feedback. We have revised the manuscript to address the concerns about methodological transparency and statistical reporting. Below we respond point by point.

read point-by-point responses

  1. Referee: Abstract and study description: The central claim of a significant discrepancy between visual cues and 'actual internal emotional states' (including the poker face effect and superior reliability of linguistic analysis) depends on an independent ground-truth measure of participants' emotions. No protocol is described for obtaining this measure (e.g., self-report timing, validation against physiological markers, or controls for demand characteristics and social-desirability bias), which is load-bearing for the reliability comparison.

    Authors: We agree that the ground-truth protocol requires explicit description. Internal emotional states were assessed via immediate post-turn self-report questionnaires (7-point Likert scales for valence, arousal, and discrete emotions) collected after selected dialogue segments to reduce recall bias. Anonymous collection and neutral instructions were used to mitigate social-desirability effects. No physiological sensors were employed to preserve ecological validity in a natural setting. The revised User Study section now details the exact questionnaire, timing, and bias controls. revision: yes

  2. Referee: User study section: The findings on emotion elicitation and disengagement are presented without reported statistical tests, effect sizes, error bars, baseline comparisons, or justification for the sample size of 20, making it difficult to evaluate whether the observed patterns support the stated conclusions about modality reliability and proactivity effects.

    Authors: We acknowledge the omission of quantitative statistical support. The revised manuscript adds non-parametric tests (Wilcoxon signed-rank) comparing modality accuracy against self-reports, with effect sizes and confidence intervals. Figures now include error bars. A power analysis and reference to comparable HCI studies justify n=20. A baseline condition with non-proactive dialogues has been added for disengagement comparisons. revision: yes

Circularity Check

0 steps flagged

Empirical user study with direct observations; no derivations or self-referential reductions

full rationale

The paper describes a user study involving 20 participants in unscripted dialogues with a proactive conversational agent, reporting observed discrepancies between facial expressions and internal states as well as the relative reliability of linguistic analysis. These are presented as empirical findings from interaction data without any equations, fitted parameters, predictions that reduce to inputs by construction, or load-bearing self-citations. The central claims rest on direct measurement of user behavior rather than any closed logical loop, rendering the work self-contained against external benchmarks of conversational interaction.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claims rest on standard user-study assumptions about accurate self-reporting of emotions and generalizability from a small sample; no free parameters or invented entities are introduced.

axioms (1)
  • domain assumption Participants' self-reported or post-interaction measures accurately reflect their true internal emotional states during the dialogues.
    This premise is required to claim a discrepancy between visual cues and 'actual' emotions.

pith-pipeline@v0.9.0 · 5760 in / 1384 out tokens · 48694 ms · 2026-05-21T09:49:30.273040+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

64 extracted references · 64 canonical work pages

  1. [1]

    Feng, Shutong and Lubis, Nurul and Geishauser, Christian and Lin, Hsien-chin and Heck, Michael and van Niekerk, Carel and Gasic, Milica , year = 2022, month = jun, booktitle =

    work page 2022

  2. [2]

    Knowledge-Based Systems , volume = 248, pages = 108861, doi =

    Emotion recognition in conversations with emotion shift detection based on multi-task learning , author =. Knowledge-Based Systems , volume = 248, pages = 108861, doi =

    work page

  3. [3]

    Computational Intelligence and Neuroscience , volume = 2022, number = 1, pages = 8032673, doi =

    Deploying Machine Learning Techniques for Human Emotion Detection , author =. Computational Intelligence and Neuroscience , volume = 2022, number = 1, pages = 8032673, doi =

    work page 2022

  4. [4]

    IEEE Access , volume = 6, number =

    Agent-Based Simulation of Smart Beds With Internet-of-Things for Exploring Big Data Analytics , author =. IEEE Access , volume = 6, number =

    work page

  5. [5]

    2021 IEEE Automatic Speech Recognition and Understanding Workshop (ASRU) , volume =

    Multimodal Emotion Recognition with High-Level Speech and Text Features , author =. 2021 IEEE Automatic Speech Recognition and Understanding Workshop (ASRU) , volume =

    work page 2021

  6. [6]

    Christ, Lukas and Amiriparian, Shahin and Baird, Alice and Kathan, Alexander and M\". The. Proceedings of the 4th on Multimodal Sentiment Analysis Challenge and Workshop: Mimicked Emotions, Humour and Personalisation , location =

    work page

  7. [7]

    IEEE Transactions on Affective Computing , volume = 14, number = 3, pages =

    Effects of Physiological Signals in Different Types of Multimodal Sentiment Estimation , author =. IEEE Transactions on Affective Computing , volume = 14, number = 3, pages =

    work page

  8. [8]

    Affect control processes:

    Jesse Hoey and Tobias Schröder and Areej Alhothali , year = 2016, journal =. Affect control processes:. doi:10.1016/j.artint.2015.09.004 , issn =

    work page doi:10.1016/j.artint.2015.09.004 2016

  9. [9]

    Emotions in Socio-cultural Interactive

    Malhotra, Aarti and Hoey, Jesse , year = 2021, booktitle =. Emotions in Socio-cultural Interactive

    work page 2021

  10. [10]

    Proceedings of the 10th International Conference on Human-Agent Interaction , location =

    Effect of Group Identity on Emotional Contagion in Dyadic Human Agent Interaction , author =. Proceedings of the 10th International Conference on Human-Agent Interaction , location =

    work page

  11. [11]

    Applied Sciences , volume = 13, number = 2, doi =

    Prediction of Emotional Empathy in Intelligent Agents to Facilitate Precise Social Interaction , author =. Applied Sciences , volume = 13, number = 2, doi =

    work page

  12. [12]

    Developing emotionally intelligent virtual social agents , author =

    work page

  13. [13]

    IEEE Transactions on Computational Social Systems , volume = 7, number = 1, pages =

    Toward Artificial Emotional Intelligence for Cooperative Social Human–Machine Interaction , author =. IEEE Transactions on Computational Social Systems , volume = 7, number = 1, pages =

    work page

  14. [14]

    2019 International Conference on Multimodal Interaction , location =

    Multimodal Learning for Identifying Opportunities for Empathetic Responses , author =. 2019 International Conference on Multimodal Interaction , location =

    work page 2019

  15. [15]

    Computational Social Networks: Tools, Perspectives and Applications , publisher =

    Real Emotions for Simulated Social Networks , author =. Computational Social Networks: Tools, Perspectives and Applications , publisher =. doi:10.1007/978-1-4471-4048-1\_16 , isbn =

    work page doi:10.1007/978-1-4471-4048-1

  16. [16]

    IEEE Transactions on Cognitive and Developmental Systems , volume = 10, number = 4, pages =

    The Perception of Emotion in Artificial Agents , author =. IEEE Transactions on Cognitive and Developmental Systems , volume = 10, number = 4, pages =

    work page

  17. [17]

    Signal Processing: Image Communication , volume = 84, pages = 115831, doi =

    Human emotion recognition by optimally fusing facial expression and speech feature , author =. Signal Processing: Image Communication , volume = 84, pages = 115831, doi =

    work page

  18. [18]

    International Journal on Recent and Innovation Trends in Computing and Communication , volume = 11, number = 1, pages =

    Smart multi-model emotion recognition system with deep learning , author =. International Journal on Recent and Innovation Trends in Computing and Communication , volume = 11, number = 1, pages =

    work page

  19. [19]

    Deng, Jia and Pang, Gaoyang and Zhang, Zhiyu and Pang, Zhibo and Yang, Huayong and Yang, Geng , year = 2019, journal =

    work page 2019

  20. [20]

    IEEE Transactions on Affective Computing , volume = 13, number = 4, pages =

    Semantic-Rich Facial Emotional Expression Recognition , author =. IEEE Transactions on Affective Computing , volume = 13, number = 4, pages =

    work page

  21. [21]

    and Alshazly, Hammam , year = 2022, journal =

    Dar, Tarim and Javed, Ali and Bourouis, Sami and Hussein, Hany S. and Alshazly, Hammam , year = 2022, journal =

    work page 2022

  22. [22]

    Emotion detection using facial expressions-

    Rani, Jyoti and Garg, Kanwal , year = 2014, journal =. Emotion detection using facial expressions-

    work page 2014

  23. [23]

    A Survey of

    Dalvi, Chirag and Rathod, Manish and Patil, Shruti and Gite, Shilpa and Kotecha, Ketan , year = 2021, journal =. A Survey of

    work page 2021

  24. [24]

    The Visual Computer , volume = 38, number = 3, pages =

    Dual integrated convolutional neural network for real-time facial expression recognition in the wild , author =. The Visual Computer , volume = 38, number = 3, pages =. doi:10.1007/s00371-021-02069-7 , issn =

    work page doi:10.1007/s00371-021-02069-7

  25. [25]

    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops , pages =

    Video Representation Learning for Conversational Facial Expression Recognition Guided by Multiple View Reconstruction , author =. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops , pages =

    work page

  26. [26]

    doi:10.3390/s24185868 , issn =

    Yang, Xincheng and Lan, Zhenping and Wang, Nan and Li, Jiansong and Wang, Yuheng and Meng, Yuwei , year = 2024, journal =. doi:10.3390/s24185868 , issn =

    work page doi:10.3390/s24185868 2024

  27. [27]

    Facial emotion recognition:

    Kaur, Manmeet and Kumar, Munish , year = 2024, journal =. Facial emotion recognition:

    work page 2024

  28. [28]

    Big Data & Society , volume = 9, number = 2, pages = 20539517221129549, doi =

    The unbearable (technical) unreliability of automated facial emotion recognition , author =. Big Data & Society , volume = 9, number = 2, pages = 20539517221129549, doi =

    work page

  29. [29]

    2408.15777 , archiveprefix =

    A Survey on Facial Expression Recognition of Static and Dynamic Emotions , author =. 2408.15777 , archiveprefix =

    work page arXiv

  30. [30]

    Sensors , volume = 19, number = 8, doi =

    A Review on Automatic Facial Expression Recognition Systems Assisted by Multimodal Sensor Data , author =. Sensors , volume = 19, number = 8, doi =

    work page

  31. [31]

    and Patil, Shruti G

    Kusal, Sheetal D. and Patil, Shruti G. and Choudrie, Jyoti and Kotecha, Ketan V. , year = 2024, month = aug, journal =. Understanding the Performance of. doi:10.1145/3643133 , issn =

    work page doi:10.1145/3643133 2024

  32. [32]

    A Review on Text-Based Emotion Detection --

    Sheetal Kusal and Shruti Patil and Jyoti Choudrie and Ketan Kotecha and Deepali Vora and Ilias Pappas , year = 2022, eprint =. A Review on Text-Based Emotion Detection --

    work page 2022

  33. [33]

    doi:10.3390/bdcc5030043 , issn =

    Kusal, Sheetal and Patil, Shruti and Kotecha, Ketan and Aluvalu, Rajanikanth and Varadarajan, Vijayakumar , year = 2021, journal =. doi:10.3390/bdcc5030043 , issn =

    work page doi:10.3390/bdcc5030043 2021

  34. [34]

    Proceedings of the 2016 conference on empirical methods in natural language processing , pages =

    Real-time speech emotion and sentiment recognition for interactive dialogue systems , author =. Proceedings of the 2016 conference on empirical methods in natural language processing , pages =

    work page 2016

  35. [35]

    Frontiers in Psychology , volume =

    Detection of emotion by text analysis using machine learning , author =. Frontiers in Psychology , volume =. doi:10.3389/fpsyg.2023.1190326 , issn =

    work page doi:10.3389/fpsyg.2023.1190326 2023

  36. [36]

    Mahmudul and Jiyad, Zakaria Masud and Mridha, M

    Maruf, Abdullah Al and Khanam, Fahima and Haque, Md. Mahmudul and Jiyad, Zakaria Masud and Mridha, M. F. and Aung, Zeyar , year = 2024, journal =. Challenges and Opportunities of Text-Based Emotion Detection:

    work page 2024

  37. [37]

    ACM Trans

    Personality-affected Emotion Generation in Dialog Systems , author =. ACM Trans. Inf. Syst. , publisher =. doi:10.1145/3655616 , issn =

    work page doi:10.1145/3655616

  38. [38]

    Knowledge-Based Systems , volume = 192, pages = 105319, doi =

    Human-machine dialogue modelling with the fusion of word- and sentence-level emotions , author =. Knowledge-Based Systems , volume = 192, pages = 105319, doi =

    work page

  39. [39]

    Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024) , publisher =

    Collecting Human-Agent Dialogue Dataset with Frontal Brain Signal toward Capturing Unexpressed Sentiment , author =. Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024) , publisher =

    work page 2024

  40. [40]

    2024 International Joint Conference on Neural Networks (IJCNN) , volume =

    Modeling Sentiment-Speaker-Dependency for Emotion Recognition in Conversation , author =. 2024 International Joint Conference on Neural Networks (IJCNN) , volume =

    work page 2024

  41. [41]

    O'Brien and Carrie Jun Cai and Meredith Ringel Morris and Percy Liang and Michael S

    Park, Joon Sung and O'Brien, Joseph and Cai, Carrie Jun and Morris, Meredith Ringel and Liang, Percy and Bernstein, Michael S. , year = 2023, booktitle =. Generative Agents:. doi:10.1145/3586183.3606763 , isbn = 9798400701320, articleno = 2, pages =

    work page doi:10.1145/3586183.3606763 2023

  42. [42]

    Generative agent‐based modeling:

    Ghaffarzadegan, Navid and Majumdar, Aritra and Williams, Ross and Hosseinichimeh, Niyousha , year = 2024, month = jan, journal =. Generative agent‐based modeling:. doi:10.1002/sdr.1761 , issn =

    work page doi:10.1002/sdr.1761 2024

  43. [43]

    IEEE Transactions on Affective Computing , volume = 14, number = 2, pages =

    Exploring the Contextual Factors Affecting Multimodal Emotion Recognition in Videos , author =. IEEE Transactions on Affective Computing , volume = 14, number = 2, pages =

    work page

  44. [44]

    Reciprocal adaptation measures for human-agent interaction evaluation , author =

    work page

  45. [45]

    Èlektronnoe modelirovanie , volume = 45, pages =

    Recognition of User Emotions Using Artificial Intelligence , author =. Èlektronnoe modelirovanie , volume = 45, pages =

    work page

  46. [46]

    and Valverde, Erick C

    Pascual, Alexander M. and Valverde, Erick C. and Kim, Jeong-in and Jeong, Jin-Woo and Jung, Yuchul and Kim, Sang-Ho and Lim, Wansu , year = 2022, journal =. doi:10.3390/s22239524 , issn =

    work page doi:10.3390/s22239524 2022

  47. [47]

    IEEE Access , volume = 7, number =

    Efficient Facial Expression Recognition Algorithm Based on Hierarchical Deep Neural Network Structure , author =. IEEE Access , volume = 7, number =

    work page

  48. [48]

    2408.07982 , archiveprefix =

    Toward a Dialogue System Using a Large Language Model to Recognize User Emotions with a Camera , author =. 2408.07982 , archiveprefix =

    work page arXiv

  49. [49]

    2024 International Conference on Communication, Computer Sciences and Engineering (IC3SE) , volume =

    Facial Emotion Recognition for Virtual Customer Service Agents , author =. 2024 International Conference on Communication, Computer Sciences and Engineering (IC3SE) , volume =

    work page 2024

  50. [50]

    Sensors , volume = 13, number = 11, pages =

    A Multimodal Emotion Detection System during Human–Robot Interaction , author =. Sensors , volume = 13, number = 11, pages =. doi:10.3390/s131115549 , issn =

    work page doi:10.3390/s131115549

  51. [51]

    Advances in Artificial Intelligence , publisher =

    Measuring Human Emotion in Short Documents to Improve Social Robot and Agent Interactions , author =. Advances in Artificial Intelligence , publisher =

    work page

  52. [52]

    Li, Shuzhen and Zhang, Tong and Chen, C. L. Philip , year = 2024, journal =

    work page 2024

  53. [53]

    BIO Web Conf

    Real-Time Information Access in Urban Environments:. BIO Web Conf. , volume = 86, pages =

    work page

  54. [54]

    S , year = 2019, booktitle =

    Abraham, Gilu K and Bhaskaran, Preethi and Jayanthi, V. S , year = 2019, booktitle =. Lung Nodule Classification in

    work page 2019

  55. [55]

    Neurocomputing , volume =

    Combining speech-based and linguistic classifiers to recognize emotion in user spoken utterances , author =. Neurocomputing , volume =. doi:10.1016/j.neucom.2017.01.120 , issn =

    work page doi:10.1016/j.neucom.2017.01.120 2017

  56. [56]

    The lexical fallacy in emotion research:

    Fiske, Alan Page , year = 2020, journal =. The lexical fallacy in emotion research:

    work page 2020

  57. [57]

    Gaya Morey, Francesc Xavier and Manresa-Yee, Cristina and Buades Rubio, Jose Maria , year = 2024, booktitle =. An. doi:10.1145/3657242.3658601 , isbn = 9798400717871, articleno = 19, pages =

    work page doi:10.1145/3657242.3658601 2024

  58. [58]

    Multimedia Tools and Applications , volume = 81, number = 27, pages =

    A Novel Approach to Cross dataset studies in Facial Expression Recognition , author =. Multimedia Tools and Applications , volume = 81, number = 27, pages =. doi:10.1007/s11042-022-13117-2 , issn =

    work page doi:10.1007/s11042-022-13117-2

  59. [59]

    Psychological Science in the Public Interest , volume = 20, number = 1, pages =

    Emotional Expressions Reconsidered: Challenges to Inferring Emotion From Human Facial Movements , author =. Psychological Science in the Public Interest , volume = 20, number = 1, pages =

    work page

  60. [60]

    Social Cognitive and Affective Neuroscience , volume = 8, number = 8, pages =

    ‘Put on your poker face’: neural systems supporting the anticipation for expressive suppression and cognitive reappraisal , author =. Social Cognitive and Affective Neuroscience , volume = 8, number = 8, pages =. doi:10.1093/scan/nss090 , issn =

    work page doi:10.1093/scan/nss090

  61. [61]

    A theory of psychological reactance , author =

    work page

  62. [62]

    Journal of Personality and Social Psychology , publisher =

    A circumplex model of affect , author =. Journal of Personality and Social Psychology , publisher =

    work page

  63. [63]

    Affective Computing , author =

    work page

  64. [64]

    Learning and Instruction , volume = 22, number = 2, pages =

    Dynamics of affective states during complex learning , author =. Learning and Instruction , volume = 22, number = 2, pages =. doi:10.1016/j.learninstruc.2011.10.001 , issn =

    work page doi:10.1016/j.learninstruc.2011.10.001 2011