pith. sign in

arxiv: 2601.13904 · v2 · submitted 2026-01-20 · 💻 cs.AI · cs.HC

PREFAB: PREFerence-based Affective Modeling for Low-Budget Self-Annotation

Jaeyoung Moon , Youjin Choi , Yucheon Park , David Melhart , Georgios N. Yannakakis , Kyung-Joong Kim This is my paper

Pith reviewed 2026-05-16 12:56 UTC · model grok-4.3

classification 💻 cs.AI cs.HC
keywords affective computingself-annotationpreference learningpeak-end ruleemotion modelingworkload reductionuser study
0
0 comments X p. Extension

The pith

PREFAB models full affective timelines from labels on only selected inflection segments using preference learning.

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

Self-annotation of affective states normally requires continuous labeling across an entire session, which proves time-consuming and prone to fatigue. PREFAB instead identifies key segments of affective change with a preference-learning model based on the peak-end rule and ordinal emotion representations. Annotators label only those segments while the system interpolates the rest of the timeline, aided by a preview mechanism for context. User studies show the approach reduces workload, raises annotator confidence, and maintains annotation quality at levels comparable to full labeling.

Core claim

PREFAB is a retrospective self-annotation method that employs a preference-learning model to detect relative affective changes and directs annotators to label only the selected segments, interpolating the remainder of the stimulus timeline. Grounded in the peak-end rule and ordinal emotion representations, it includes a preview mechanism for contextual cues. Evaluations indicate superior performance over baselines in capturing inflections with reduced workload and higher confidence.

What carries the argument

The preference-learning model that identifies relative affective changes from limited segment labels to enable interpolation of the full affective timeline.

If this is right

  • Annotators produce complete affective timelines with reduced cognitive and temporal demands.
  • Annotation quality matches or exceeds that of full continuous labeling.
  • Annotators gain increased confidence in their provided labels.
  • The method outperforms standard baselines in modeling accuracy for inflections.

Where Pith is reading between the lines

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

  • This sparse labeling strategy could generalize to other continuous labeling domains like physiological signal annotation.
  • Combining PREFAB with automated detection of inflections might reduce human effort even further.
  • The reliance on ordinal preferences suggests that absolute scale ratings may not always be necessary for accurate timeline reconstruction.

Load-bearing premise

The preference-learning model grounded in the peak-end rule can detect relative affective changes accurately enough from limited labels to interpolate the timeline without systematic bias.

What would settle it

A controlled comparison where the full ground-truth affective timeline is available from exhaustive labeling, and PREFAB's interpolated version shows significantly higher error rates in non-labeled segments than expected.

Figures

Figures reproduced from arXiv: 2601.13904 by David Melhart, Georgios N. Yannakakis, Jaeyoung Moon, Kyung-Joong Kim, Youjin Choi, Yucheon Park.

Figure 1
Figure 1. Figure 1: Full self-annotation imposes high cognitive workload (left). PREFAB alleviates this by (a) reconstructing expected [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Visualization of BCE and OCE probability func [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Evaluation and results of dynamic time warping [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Architecture of the PREFAB model. The model takes two consecutive segments sampled at a 1-second interval, each [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Model performance across nine games. The top pan [PITH_FULL_IMAGE:figures/full_fig_p008_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Comparison of inflection point sampling across methods. X-axis indicates timestamps and Y-axis arousal levels. Green [PITH_FULL_IMAGE:figures/full_fig_p009_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Overall Process of User Study trace and the original full-annotation ground truth was quanti￾fied using three metrics: Concordance Correlation Coefficient (CCC) [29] for absolute agreement, Spearman’s rank correlation (𝜌) [61] for trend similarity, and Dynamic Time Warping (DTW) Similarity [56] for morphological alignment. Second, we compared the tem￾poral characteristics of PREFAB’s sampling against the i… view at source ↗
Figure 8
Figure 8. Figure 8: Results of self-reported workload and confidence [PITH_FULL_IMAGE:figures/full_fig_p012_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Results of self-reported annotation result quality [PITH_FULL_IMAGE:figures/full_fig_p014_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Comparison of the baseline arousal trace and its [PITH_FULL_IMAGE:figures/full_fig_p014_10.png] view at source ↗
read the original abstract

Self-annotation is the gold standard for collecting affective state labels in affective computing. Existing methods typically rely on full annotation, requiring users to continuously label affective states across entire sessions. While this process yields fine-grained data, it is time-consuming, cognitively demanding, and prone to fatigue and errors. To address these issues, we present PREFAB, a low-budget retrospective self-annotation method that targets affective inflection regions rather than full annotation. Grounded in the peak-end rule and ordinal representations of emotion, PREFAB employs a preference-learning model to detect relative affective changes, directing annotators to label only selected segments while interpolating the remainder of the stimulus. We further introduce a preview mechanism that provides brief contextual cues to assist annotation. We evaluate PREFAB through a technical performance study and a 25-participant user study. Results show that PREFAB outperforms baselines in modeling affective inflections while mitigating workload (and conditionally mitigating temporal burden). Importantly PREFAB improves annotator confidence without degrading annotation quality.

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

1 major / 1 minor

Summary. The manuscript introduces PREFAB, a retrospective self-annotation technique for affective states that leverages the peak-end rule to identify key segments for labeling, employs a preference-learning model for interpolation of the affective timeline, and includes a preview mechanism. Through a technical performance study and a 25-participant user study, it claims to outperform baselines in modeling affective inflections, reduce workload, and improve annotator confidence without degrading quality.

Significance. If the core assumptions hold, PREFAB could enable more efficient collection of affective data in computing applications by substantially lowering annotation effort while preserving data utility, which is valuable for scaling studies in affective computing and human-computer interaction.

major comments (1)
  1. [Evaluation / Results] The central claim that annotation quality is preserved relies on the assumption that the preference model and interpolation from peak-end selected segments accurately reconstruct the full affective timeline. However, the reported user study compares PREFAB to baselines on workload and confidence but does not include a direct within-subject validation of the interpolated values against a complete per-frame ground truth annotation of the same stimuli. This leaves the quality preservation claim untested against potential systematic temporal biases.
minor comments (1)
  1. [Abstract] The abstract mentions 'outperforms baselines' and 'mitigating workload' but does not specify the exact metrics, statistical tests, or baseline definitions used in the evaluation.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their constructive and insightful review of our manuscript. We address the single major comment below, acknowledging its validity while providing the strongest honest defense of our evaluation approach and outlining the revisions we will undertake.

read point-by-point responses

  1. Referee: [Evaluation / Results] The central claim that annotation quality is preserved relies on the assumption that the preference model and interpolation from peak-end selected segments accurately reconstruct the full affective timeline. However, the reported user study compares PREFAB to baselines on workload and confidence but does not include a direct within-subject validation of the interpolated values against a complete per-frame ground truth annotation of the same stimuli. This leaves the quality preservation claim untested against potential systematic temporal biases.

    Authors: We thank the referee for highlighting this important methodological point. The technical performance study evaluates the preference-learning model's reconstruction accuracy by comparing interpolated timelines against full per-frame annotations on held-out data, providing quantitative support for the interpolation approach under the peak-end rule. The 25-participant user study was deliberately scoped to assess real-world usability (workload reduction and annotator confidence) while preserving the low-budget character of PREFAB; requiring full per-frame ground-truth annotations from the same participants would have defeated that purpose. We therefore agree that the user-study results do not furnish a direct within-subject test of reconstruction fidelity and that potential temporal biases remain unexamined in that specific setting. In the revised manuscript we will (1) clearly separate the evidence provided by the technical study from the usability findings, (2) temper the claim of “without degrading annotation quality” to reflect the available evidence, (3) add an explicit discussion of possible systematic temporal biases, and (4) list direct within-subject validation as a concrete direction for future work. These changes constitute a partial revision: the text and claims will be updated, but we cannot retroactively collect new full-annotation data for the existing user-study cohort. revision: partial

Circularity Check

0 steps flagged

No significant circularity; claims rest on external user-study validation

full rationale

The derivation applies the peak-end rule and preference learning to select segments for labeling, then interpolates the timeline. Evaluation proceeds via a technical performance study and 25-participant user study that compare PREFAB against baselines on independent metrics (workload, confidence, annotation quality). No equations reduce by construction to fitted parameters, self-definitions, or self-citation chains; the central performance claims are externally falsifiable against the study outcomes rather than tautological.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 1 invented entities

The central claim rests on the peak-end rule and ordinal emotion representations as background assumptions plus the new PREFAB procedure itself; no explicit free parameters or invented physical entities are named in the abstract.

axioms (2)
  • domain assumption The peak-end rule governs how people remember and report affective experiences
    Invoked to justify focusing annotation effort on inflection regions rather than full sessions
  • domain assumption Affective states can be represented ordinally for preference comparison
    Used to enable relative change detection via preference learning
invented entities (1)
  • PREFAB method no independent evidence
    purpose: Low-budget retrospective self-annotation targeting affective inflections
    New protocol introduced in the paper; no independent evidence provided beyond the described studies

pith-pipeline@v0.9.0 · 5491 in / 1483 out tokens · 47123 ms · 2026-05-16T12:56:45.428114+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

77 extracted references · 77 canonical work pages

  1. [1]

    Akhilesh Adithya, Snigdha Tiwari, Sougata Sen, Sandip Chakraborty, and Surjya Ghosh. 2022. OCEAN: Towards Developing an Opportunistic Continuous Emo- tion Annotation Framework. In2022 IEEE International Conference on Pervasive Computing and Communications Workshops and other Affiliated Events (PerCom Workshops). 9–12. https://doi.org/10.1109/PerComWorksho...

    work page doi:10.1109/percomworkshops53856.2022.9767359 2022

  2. [2]

    Değer Ayata, Yusuf Yaslan, and Mustafa Kamaşak. 2016. Emotion recognition via random forest and galvanic skin response: Comparison of time based feature sets, window sizes and wavelet approaches. In2016 Medical Technologies National Congress (TIPTEKNO). 1–4. https://doi.org/10.1109/TIPTEKNO.2016.7863130

    work page doi:10.1109/tiptekno.2016.7863130 2016

  3. [3]

    Bailey and Joseph A

    Brian P. Bailey and Joseph A. Konstan. 2006. On the need for attention-aware systems: Measuring effects of interruption on task performance, error rate, and affective state.Computers in Human Behavior22, 4 (2006), 685–708. https: //doi.org/10.1016/j.chb.2005.12.009 Attention aware systems

    work page doi:10.1016/j.chb.2005.12.009 2006

  4. [4]

    Swarnali Banik, Sougata Sen, Snehanshu Saha, and Surjya Ghosh. 2024. Towards Reducing Continuous Emotion Annotation Effort During Video Consumption: A Physiological Response Profiling Approach.Proc. ACM Interact. Mob. Wearable Ubiquitous Technol.8, 3, Article 91 (Sept. 2024), 32 pages. https://doi.org/10.1145/ 3678569

    work page 2024

  5. [5]

    Virginia Braun and Victoria Clarke. 2006. Using thematic analysis in psychology.Qualitative Research in Psychology3, 2 (2006), 77–101. https://doi.org/10.1191/1478088706qp063oa arXiv:https://www.tandfonline.com/doi/pdf/10.1191/1478088706qp063oa

    work page doi:10.1191/1478088706qp063oa 2006

  6. [6]

    Broderick, Joseph E

    Joan E. Broderick, Joseph E. Schwartz, Gregory Vikingstad, Michelle Pribbernow, Steven Grossman, and Arthur A. Stone. 2008. The accuracy of pain and fatigue items across different reporting periods.PAIN139, 1 (2008), 146–157. https: //doi.org/10.1016/j.pain.2008.03.024

    work page doi:10.1016/j.pain.2008.03.024 2008

  7. [7]

    Joost Broekens, Anne Pronker, and Marian Neuteboom. 2010. Real time labeling of affect in music using the affectbutton. InProceedings of the 3rd International Workshop on Affective Interaction in Natural Environments(Firenze, Italy)(AFFINE ’10). Association for Computing Machinery, New York, NY, USA, 21–26. https: //doi.org/10.1145/1877826.1877833

    work page doi:10.1145/1877826.1877833 2010

  8. [8]

    Chris Burges, Tal Shaked, Erin Renshaw, Ari Lazier, Matt Deeds, Nicole Hamilton, and Greg Hullender. 2005. Learning to rank using gradient descent. InProceedings of the 22nd international conference on Machine learning(Bonn, Germany)(ICML ’05). Association for Computing Machinery, New York, NY, USA, 89–96. https: //doi.org/10.1145/1102351.1102363

    work page doi:10.1145/1102351.1102363 2005

  9. [9]

    Elizabeth Camilleri, Georgios N Yannakakis, and Antonios Liapis. 2017. Towards general models of player affect. In2017 Seventh International Conference on Affective Computing and Intelligent Interaction (ACII). IEEE, 333–339

    work page 2017

  10. [10]

    Jingjing Chen, Xiaobin Wang, Chen Huang, Xin Hu, Xinke Shen, and Dan Zhang

    work page

  11. [11]

    https://doi.org/10.1038/s41597-023-02650-w

    A Large Finer-grained Affective Computing EEG Dataset.Scientific Data 10, 1 (25 Oct 2023), 740. https://doi.org/10.1038/s41597-023-02650-w

    work page doi:10.1038/s41597-023-02650-w 2023

  12. [12]

    Youjin Choi, JooYeong Kim, and Jin-Hyuk Hong. 2022. Immersion Measurement in Watching Videos Using Eye-tracking Data.IEEE Transactions on Affective Computing13, 4 (2022), 1759–1770. https://doi.org/10.1109/TAFFC.2022.3209311

    work page doi:10.1109/taffc.2022.3209311 2022

  13. [13]

    Roddy Cowie, Ellen Douglas-Cowie, Susie Savvidou, Edelle McMahon, Martin Sawey, and Marc Schröder. 2000. Feeltrace: An instrument for recording perceived emotion in real time. InProceedings of the ISCA Workshop on Speech and Emotion, Vol. 1

    work page 2000

  14. [14]

    Roddy Cowie, Martin Sawey, Cian Doherty, Javier Jaimovich, Cavan Fyans, and Paul Stapleton. 2013. Gtrace: General Trace Program Compatible with Emo- tionML. In2013 Humaine Association Conference on Affective Computing and Intelligent Interaction. 709–710. https://doi.org/10.1109/ACII.2013.126

    work page doi:10.1109/acii.2013.126 2013

  15. [15]

    MIHALY CSIKSZENTMIHALYI and REED LARSON. 1987. Validity and Reliability of the Experience-Sampling Method.The Journal of Nervous and Mental Disease 175, 9 (1987). https://journals.lww.com/jonmd/fulltext/1987/09000/validity_and_ reliability_of_the.4.aspx

    work page 1987

  16. [16]

    Bridgid Finn. 2010. Ending on a high note: Adding a better end to effortful study. Journal of Experimental Psychology: Learning, Memory, and Cognition36, 6 (2010), 1548

    work page 2010

  17. [17]

    Barbara L Fredrickson and Daniel Kahneman. 1993. Duration neglect in ret- rospective evaluations of affective episodes.Journal of personality and social psychology65, 1 (1993), 45

    work page 1993

  18. [18]

    Siddhant Garg and Yingyu Liang. 2020. Functional regularization for representa- tion learning: A unified theoretical perspective.Advances in Neural Information Processing Systems33 (2020), 17187–17199

    work page 2020

  19. [19]

    Jeffrey M Girard. 2014. CARMA: Software for continuous affect rating and media annotation.Journal of open research software2, 1 (2014), e5

    work page 2014

  20. [20]

    Jeffrey M Girard and Aidan G C. Wright. 2018. DARMA: Software for dual axis rating and media annotation.Behavior research methods50 (2018), 902–909

    work page 2018

  21. [21]

    Shalom Greene, Himanshu Thapliyal, and Allison Caban-Holt. 2016. A Survey of Affective Computing for Stress Detection: Evaluating technologies in stress detection for better health.IEEE Consumer Electronics Magazine5, 4 (2016), 44–56. https://doi.org/10.1109/MCE.2016.2590178

    work page doi:10.1109/mce.2016.2590178 2016

  22. [22]

    Kyeong-Soo Han, Han kyu Lee, Youngho Jeong, and Youn-Seon Jang. 2013. Improvement of viewers’ preference inference by applying the peak-end rule. In2013 International Conference on ICT Convergence (ICTC). 792–797. https: //doi.org/10.1109/ICTC.2013.6675481

    work page doi:10.1109/ictc.2013.6675481 2013

  23. [23]

    Sepp Hochreiter and Jürgen Schmidhuber. 1997. Long Short-Term Memory. Neural Computation9, 8 (1997), 1735–1780. https://doi.org/10.1162/neco.1997.9. 8.1735

    work page doi:10.1162/neco.1997.9 1997

  24. [24]

    Alexander Hoelzemann and Kristof Van Laerhoven. 2024. A matter of annotation: an empirical study on in situ and self-recall activity annotations from wearable sensors.Frontiers in Computer Science6 (2024), 1379788

    work page 2024

  25. [25]

    Ziniu Hu, Zhe Zhao, Xinyang Yi, Tiansheng Yao, Lichan Hong, Yizhou Sun, and Ed Chi. 2022. Improving Multi-Task Generalization via Regularizing Spurious Correlation. InAdvances in Neural Information Processing Systems, S. Koyejo, S. Mohamed, A. Agarwal, D. Belgrave, K. Cho, and A. Oh (Eds.), Vol. 35. Curran Associates, Inc., 11450–11466. https://proceeding...

    work page 2022

  26. [26]

    Patrícia Augustin Jaques, Matheus Lehmann, and Sylvie Pesty. 2009. Evaluating the affective tactics of an emotional pedagogical agent. InProceedings of the 2009 ACM Symposium on Applied Computing(Honolulu, Hawaii)(SAC ’09). Association for Computing Machinery, New York, NY, USA, 104–109. https://doi.org/10. 1145/1529282.1529304

    work page arXiv 2009

  27. [27]

    Soheil Khorram, Melvin G McInnis, and Emily Mower Provost. 2019. Jointly aligning and predicting continuous emotion annotations.IEEE Transactions on Affective Computing12, 4 (2019), 1069–1083

    work page 2019

  28. [28]

    Gregory Koch, Richard Zemel, Ruslan Salakhutdinov, et al. 2015. Siamese neural networks for one-shot image recognition. InICML deep learning workshop, Vol. 2. Lille, 1–30

    work page 2015

  29. [29]

    Kurby and Jeffrey M

    Christopher A. Kurby and Jeffrey M. Zacks. 2008. Segmentation in the perception and memory of events.Trends in Cognitive Sciences12, 2 (2008), 72–79. https: //doi.org/10.1016/j.tics.2007.11.004

    work page doi:10.1016/j.tics.2007.11.004 2008

  30. [30]

    I Lawrence and Kuei Lin. 1989. A concordance correlation coefficient to evaluate reproducibility.Biometrics(1989), 255–268. PREFAB: PREFerence-based Affective Modeling for Low-Budget Self-Annotation CHI ’26, April 13–17, 2026, Barcelona, Spain

    work page 1989

  31. [31]

    Zongxia Li, Xiyang Wu, Hongyang Du, Fuxiao Liu, Huy Nghiem, and Guangyao Shi. 2025. A Survey of State of the Art Large Vision Language Models: Bench- mark Evaluations and Challenges. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops. 1587–1606

    work page 2025

  32. [32]

    Phil Lopes, Georgios N Yannakakis, and Antonios Liapis. 2017. Ranktrace: Relative and unbounded affect annotation. In2017 Seventh International Conference on Affective Computing and Intelligent Interaction (ACII). IEEE, 158–163

    work page 2017

  33. [33]

    Yiyue Luo, Yunzhu Li, Pratyusha Sharma, Wan Shou, Kui Wu, Michael Foshey, Beichen Li, Tomás Palacios, Antonio Torralba, and Wojciech Matusik. 2021. Learn- ing human–environment interactions using conformal tactile textiles.Nature Electronics4, 3 (2021), 193–201

    work page 2021

  34. [34]

    Yiyue Luo, Kui Wu, Tomás Palacios, and Wojciech Matusik. 2021. KnitUI: Fab- ricating Interactive and Sensing Textiles with Machine Knitting. InProceedings of the 2021 CHI Conference on Human Factors in Computing Systems(Yokohama, Japan)(CHI ’21). Association for Computing Machinery, New York, NY, USA, Article 668, 12 pages. https://doi.org/10.1145/3411764.3445780

    work page doi:10.1145/3411764.3445780 2021

  35. [35]

    Clare Lyle, Mark Rowland, Georg Ostrovski, and Will Dabney. 2021. On the effect of auxiliary tasks on representation dynamics. InInternational Conference on Artificial Intelligence and Statistics. PMLR, 1–9

    work page 2021

  36. [36]

    Yannakakis

    Konstantinos Makantasis, Antonios Liapis, and Georgios N. Yannakakis. 2023. The Pixels and Sounds of Emotion: General-Purpose Representations of Arousal in Games.IEEE Transactions on Affective Computing14, 1 (2023), 680–693. https: //doi.org/10.1109/TAFFC.2021.3060877

    work page doi:10.1109/taffc.2021.3060877 2023

  37. [37]

    Soroosh Mariooryad and Carlos Busso. 2013. Analysis and Compensation of the Reaction Lag of Evaluators in Continuous Emotional Annotations. In2013 Humaine Association Conference on Affective Computing and Intelligent Interaction. 85–90. https://doi.org/10.1109/ACII.2013.21

    work page doi:10.1109/acii.2013.21 2013

  38. [38]

    Soroosh Mariooryad and Carlos Busso. 2015. Correcting Time-Continuous Emo- tional Labels by Modeling the Reaction Lag of Evaluators.IEEE Transactions on Affective Computing6, 2 (2015), 97–108. https://doi.org/10.1109/TAFFC.2014. 2334294

    work page doi:10.1109/taffc.2014 2015

  39. [39]

    Gloria Mark, Daniela Gudith, and Ulrich Klocke. 2008. The cost of interrupted work: more speed and stress. InProceedings of the SIGCHI Conference on Human Factors in Computing Systems(Florence, Italy)(CHI ’08). Association for Comput- ing Machinery, New York, NY, USA, 107–110. https://doi.org/10.1145/1357054. 1357072

    work page doi:10.1145/1357054 2008

  40. [40]

    Martínez, Georgios N

    Héctor P. Martínez, Georgios N. Yannakakis, and John Hallam. 2014. Don’t Classify Ratings of Affect; Rank Them!IEEE Transactions on Affective Computing 5, 3 (2014), 314–326. https://doi.org/10.1109/TAFFC.2014.2352268

    work page doi:10.1109/taffc.2014.2352268 2014

  41. [41]

    David Melhart, Antonios Liapis, and Georgios N Yannakakis. 2019. PAGAN: Video affect annotation made easy. In2019 8th international conference on affective computing and intelligent interaction (ACII). IEEE, 130–136

    work page 2019

  42. [42]

    David Melhart, Antonios Liapis, and Georgios N Yannakakis. 2022. The arousal video game annotation (AGAIN) dataset.IEEE Transactions on Affective Computing 13, 4 (2022), 2171–2184

    work page 2022

  43. [43]

    David Melhart, Georgios N Yannakakis, and Antonios Liapis. 2020. I feel i feel you: A theory of mind experiment in games.KI-Künstliche Intelligenz34, 1 (2020), 45–55

    work page 2020

  44. [44]

    Metacritic. [n. d.].Metacritic. https://www.metacritic.com/game

    work page

  45. [45]

    Angeliki Metallinou and Shrikanth Narayanan. 2013. Annotation and processing of continuous emotional attributes: Challenges and opportunities. In2013 10th IEEE International Conference and Workshops on Automatic Face and Gesture Recognition (FG). 1–8. https://doi.org/10.1109/FG.2013.6553804

    work page doi:10.1109/fg.2013.6553804 2013

  46. [46]

    JaeYoung Moon, EunHye Cho, Yeabon Jo, KyungJoong Kim, and Eunsung Song

    work page

  47. [47]

    https://doi.org/10.1109/TG.2024.3424459

    Investigating the Effect of Emotional Matching Between Game and Back- ground Music on Game Experience in a Valence–Arousal Space.IEEE Transactions on Games17, 2 (2025), 282–295. https://doi.org/10.1109/TG.2024.3424459

    work page doi:10.1109/tg.2024.3424459 2025

  48. [48]

    JaeYoung Moon, YouJin Choi, TaeHwa Park, JunDoo Choi, Jin-Hyuk Hong, and Kyung-Joong Kim. 2022. Diversifying dynamic difficulty adjustment agent by integrating player state models into Monte-Carlo tree search.Expert Systems with Applications205 (2022), 117677. https://doi.org/10.1016/j.eswa.2022.117677

    work page doi:10.1016/j.eswa.2022.117677 2022

  49. [49]

    Masaki Omata, Daisuke Kanuka, Xiaoyang Mao, and Atsumi Imamiya. 2012. Af- fective Doodle: a painting tool reflecting user emotion. InProceedings of the ACM Symposium on Applied Perception(Los Angeles, California)(SAP ’12). Association for Computing Machinery, New York, NY, USA, 120. https://doi.org/10.1145/ 2338676.2338702

    work page arXiv 2012

  50. [50]

    Nathalie Pattyn, Xavier Neyt, David Henderickx, and Eric Soetens. 2008. Psy- chophysiological investigation of vigilance decrement: Boredom or cognitive fatigue?Physiology & Behavior93, 1 (2008), 369–378. https://doi.org/10.1016/j. physbeh.2007.09.016

    work page doi:10.1016/j 2008

  51. [51]

    Fabian Pedregosa, Francis Bach, and Alexandre Gramfort. 2017. On the Consis- tency of Ordinal Regression Methods.Journal of Machine Learning Research18, 55 (2017), 1–35. http://jmlr.org/papers/v18/15-495.html

    work page 2017

  52. [52]

    Guanxiong Pei, Haiying Li, Yandi Lu, Yanlei Wang, Shizhen Hua, and Taihao Li

    work page

  53. [53]

    Intelligent Computing3 (2024), 0076

    Affective Computing: Recent Advances, Challenges, and Future Trends. Intelligent Computing3 (2024), 0076. https://doi.org/10.34133/icomputing.0076 arXiv:https://spj.science.org/doi/pdf/10.34133/icomputing.0076

    work page doi:10.34133/icomputing.0076 2024

  54. [54]

    Ethan Perez, Florian Strub, Harm de Vries, Vincent Dumoulin, and Aaron Courville. 2018. FiLM: Visual Reasoning with a General Conditioning Layer. Proceedings of the AAAI Conference on Artificial Intelligence32, 1 (Apr. 2018). https://doi.org/10.1609/aaai.v32i1.11671

    work page doi:10.1609/aaai.v32i1.11671 2018

  55. [55]

    Rosalind W. Picard. 1997.Affective computing. MIT Press, Cambridge, MA, USA

    work page 1997

  56. [56]

    Yan- nakakis

    Kosmas Pinitas, Konstantinos Makantasis, Antonios Liapis, and Georgios N. Yan- nakakis. 2022. RankNEAT: outperforming stochastic gradient search in preference learning tasks. InProceedings of the Genetic and Evolutionary Computation Confer- ence(Boston, Massachusetts)(GECCO ’22). Association for Computing Machinery, New York, NY, USA, 1084–1092. https://d...

    work page doi:10.1145/3512290.3528744 2022

  57. [57]

    Dominic Potts, Miloni Gada, Aastha Gupta, Kavya Goel, Klaus Philipp Krzok, Genevieve Pate, Joseph Hartley, Mark Weston-Arnold, Jakob Aylott, Christopher Clarke, Crescent Jicol, and Christof Lutteroth. 2025. RetroSketch: A Retrospective Method for Measuring Emotions and Presence in Virtual Reality. InProceedings of the 2025 CHI Conference on Human Factors ...

    work page doi:10.1145/3706598.3713957 2025

  58. [58]

    Michael D Robinson and Gerald L Clore. 2002. Belief and feeling: evidence for an accessibility model of emotional self-report.Psychological bulletin128, 6 (2002), 934

    work page 2002

  59. [59]

    Sakoe and S

    H. Sakoe and S. Chiba. 1978. Dynamic programming algorithm optimization for spoken word recognition.IEEE Transactions on Acoustics, Speech, and Signal Processing26, 1 (1978), 43–49. https://doi.org/10.1109/TASSP.1978.1163055

    work page doi:10.1109/tassp.1978.1163055 1978

  60. [60]

    Julian Scharbert, Katharina Utesch, Thomas Reiter, Julian ter Horst, Maarten van Zalk, Mitja D Back, and Richard Rau. 2025. If you were happy and you know it, clap your hands! Testing the peak-end rule for retrospec- tive judgments of well-being in everyday life.European Journal of Per- sonality39, 1 (2025), 55–69. https://doi.org/10.1177/0890207024123596...

    work page doi:10.1177/08902070241235969 2025

  61. [61]

    van den Broek, Alin Albu-Schaeffer, and Friedhelm Schwenker

    Karan Sharma, Claudio Castellini, Egon L. van den Broek, Alin Albu-Schaeffer, and Friedhelm Schwenker. 2019. A dataset of continuous affect annotations and physiological signals for emotion analysis.Scientific Data6, 1 (09 Oct 2019), 196. https://doi.org/10.1038/s41597-019-0209-0

    work page doi:10.1038/s41597-019-0209-0 2019

  62. [62]

    Saul Shiffman, Arthur A Stone, and Michael R Hufford. 2008. Ecological momen- tary assessment.Annu. Rev. Clin. Psychol.4, 1 (2008), 1–32

    work page 2008

  63. [63]

    Jianping Song, Tianran Tang, and Guosheng Hu. 2023. The duration threshold of video content observation: An experimental investigation of visual perception efficiency.Computer Science and Information Systems20, 2 (2023), 879–892

    work page 2023

  64. [64]

    Spearman

    C. Spearman. 1904. The Proof and Measurement of Association between Two Things.The American Journal of Psychology15, 1 (1904), 72–101. http://www. jstor.org/stable/1412159

    work page arXiv 1904

  65. [65]

    Jianhua Tao and Tieniu Tan. 2005. Affective Computing: A Review. InAffective Computing and Intelligent Interaction, Jianhua Tao, Tieniu Tan, and Rosalind W. Picard (Eds.). Springer Berlin Heidelberg, Berlin, Heidelberg, 981–995

    work page 2005

  66. [66]

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need.Advances in neural information processing systems30 (2017)

    work page 2017

  67. [67]

    Virtanen, R

    Pauli Virtanen, Ralf Gommers, Travis E. Oliphant, Matt Haberland, Tyler Reddy, David Cournapeau, Evgeni Burovski, Pearu Peterson, Warren Weckesser, Jonathan Bright, Stéfan J. van der Walt, Matthew Brett, Joshua Wilson, K. Jar- rod Millman, Nikolay Mayorov, Andrew R. J. Nelson, Eric Jones, Robert Kern, Eric Larson, C J Carey, İlhan Polat, Yu Feng, Eric W. ...

    work page doi:10.1038/s41592-019-0686-2 2020

  68. [68]

    Shuo Yan, Gangyi Ding, Hongsong Li, Ningxiao Sun, Zheng Guan, Yufeng Wu, Longfei Zhang, and Tianyu Huang. 2017. Exploring Audience Response in Performing Arts with a Brain-Adaptive Digital Performance System.ACM Trans. Interact. Intell. Syst.7, 4, Article 16 (Dec. 2017), 28 pages. https://doi.org/10.1145/ 3009974

    work page 2017

  69. [69]

    Yannakakis

    Georgios N. Yannakakis. 2009. Preference learning for affective modeling. In2009 3rd International Conference on Affective Computing and Intelligent Interaction and Workshops. 1–6. https://doi.org/10.1109/ACII.2009.5349491

    work page doi:10.1109/acii.2009.5349491 2009

  70. [70]

    Georgios N Yannakakis, Roddy Cowie, and Carlos Busso. 2018. The ordinal nature of emotions: An emerging approach.IEEE Transactions on Affective Computing 12, 1 (2018), 16–35

    work page 2018

  71. [71]

    Yannakakis, Manolis Maragoudakis, and John Hallam

    Georgios N. Yannakakis, Manolis Maragoudakis, and John Hallam. 2009. Prefer- ence Learning for Cognitive Modeling: A Case Study on Entertainment Prefer- ences.IEEE Transactions on Systems, Man, and Cybernetics - Part A: Systems and Humans39, 6 (2009), 1165–1175. https://doi.org/10.1109/TSMCA.2009.2028152

    work page doi:10.1109/tsmca.2009.2028152 2009

  72. [72]

    Yannakakis and Héctor P

    Georgios N. Yannakakis and Héctor P. Martínez. 2015. Grounding truth via ordinal annotation. In2015 International Conference on Affective Computing and Intelligent Interaction (ACII). 574–580. https://doi.org/10.1109/ACII.2015.7344627

    work page doi:10.1109/acii.2015.7344627 2015

  73. [73]

    Bremner, Louise A

    Georgios N. Yannakakis and David Melhart. 2023. Affective Game Computing: A Survey.Proc. IEEE111, 10 (2023), 1423–1444. https://doi.org/10.1109/JPROC. CHI ’26, April 13–17, 2026, Barcelona, Spain Moon et al. 2023.3315689

    work page doi:10.1109/jproc 2023

  74. [74]

    Jeffrey M Zacks, Nicole K Speer, Khena M Swallow, Todd S Braver, and Jeremy R Reynolds. 2007. Event perception: a mind-brain perspective.Psychological bulletin 133, 2 (2007), 273

    work page 2007

  75. [75]

    Emma Yann Zhang, Zhigeng Pan, and Adrian David Cheok. 2025. Emotion Recognition Using Affective Touch: A Survey.IEEE Transactions on Affective Computing(2025), 1–20. https://doi.org/10.1109/TAFFC.2025.3592197

    work page doi:10.1109/taffc.2025.3592197 2025

  76. [76]

    Tianyi Zhang, Abdallah El Ali, Chen Wang, Alan Hanjalic, and Pablo Cesar. 2020. Rcea: Real-time, continuous emotion annotation for collecting precise mobile video ground truth labels. InProceedings of the 2020 CHI Conference on Human Factors in Computing Systems. 1–15

    work page 2020

  77. [77]

    Zixin Zhang, Fan Qi, Shuai Li, and Changsheng Xu. 2023. AffectFAL: Fed- erated Active Affective Computing with Non-IID Data. InProceedings of the 31st ACM International Conference on Multimedia(Ottawa ON, Canada)(MM ’23). Association for Computing Machinery, New York, NY, USA, 871–882. https://doi.org/10.1145/3581783.3612442 PREFAB: PREFerence-based Affec...

    work page doi:10.1145/3581783.3612442 2023