Pith. sign in

REVIEW 3 major objections 5 minor 49 references

Emotions as Ambiguity-aware Ordinal Representations

T0 review · 3 major / 5 minor · reviewed 2026-08-05 · deepseek-v4-flash

Pith's one-line read Continuous emotion traces can be modeled by their rate of change, and this ordinal representation outperforms interval-based ambiguity models on unbounded labels.

desk verdict A clean idea for treating ambiguity as dynamic, but the main comparison is between different prediction targets, so the headline claim about ordinal superiority isn't established. read the letter →

arxiv 2508.19193 v2 pith:DBLLXLEO submitted 2025-08-26 cs.LG cs.AI

classification cs.LGcs.AI
keywords continuousemotionrecognitionordinalrepresentationinter-raterambiguitytemporaldynamicsengagementpredictionrateofchangeaffectivecomputingunboundedlabels
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 argues that the ambiguity of emotion annotations should be modeled through how ratings change over time, not through their absolute values at each moment. It introduces two ordinal representations: one tracks each annotator's gradient and fits a distribution over those gradients, the other tracks the gradients of the group's mean and standard deviation. On the unbounded engagement labels of GameVibe, the group ordinal representation gives the highest concordance and directional-agreement scores; on the bounded arousal and valence labels of RECOLA, ordinal representations improve directional agreement even though interval representations keep the best absolute-value fit. The takeaway is that for predicting the direction of emotional change, relative change carries more signal than absolute magnitude.

What carries the argument

The central object is the gradient, the rate of change of an annotation trace over time, used as the unit of emotional meaning. The individual representation computes each annotator's gradient at each time step and fits a probability distribution over those gradients, so the mean captures agreed-upon direction and the standard deviation captures disagreement about that direction. The group representation instead computes the gradients of the interval representation's own parameters, dμ/dt and dσ/dt, directly tracking how the group's central emotion and its ambiguity are moving. The gradient transform is what converts static ambiguity into dynamic, ordinal information, and it is what the pape

What would settle it

If the unbounded GameVibe traces were re-analyzed after removing the annotators' progressive divergence, for example by normalizing each trace to zero mean or detrending, and the group ordinal representation no longer beat the interval representation on SDA and CCC, then the claimed advantage would be an artifact of global drift rather than of ordinal structure.

Watch

Extended reading notes

Core claim

The paper's central claim is that a continuous emotion trace is better represented, for recognition purposes, by the rate of change of its annotations than by the absolute interval values, and that ambiguity should be encoded as the spread of those changes. Concretely, it defines two ambiguity-aware ordinal representations: the individual representation fits a probability distribution over the temporal gradients of each annotator's trace, while the group representation takes the temporal gradients of the group's mean and standard deviation. Tested in an LSTM-based continuous emotion recognition system, the group representation outperforms interval and individual representations on GameVibe's

Load-bearing premise

The load-bearing premise is that people agree more on how an emotion changes over time than on its absolute magnitude, and that this relative agreement is the signal worth modeling; the paper takes this premise from prior work rather than testing it directly.

Editorial extensions

If this is right

  • For unbounded continuous emotion labels, ordinal representations should be a strong default target: group-level gradient modeling yielded the top CCC and SDA scores on GameVibe engagement.
  • For bounded traces, ordinal representations are complementary to interval ones: they improve SDA relative to interval representations even when absolute CCC drops.
  • The group representation is the more reliable variant: aggregating trends across annotators smoothed individual inconsistencies and consistently outperformed per-annotator gradient modeling.
  • Evaluation of ordinal emotion models should include SDA, since it directly measures agreement in directional trend rather than magnitude.
  • Existing interval-based ambiguity models may under-use the temporal information in rater disagreement.

Reading between the lines

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

  • The gradient-based representation is not emotion-specific: any continuous subjective rating with multiple annotators, such as pain, fatigue, or product satisfaction, could be recast in the same ordinal form and might show similar directional-agreement gains.
  • The paper leaves implicit that the advantage of the group representation on sigma may hinge on the unbounded traces' tendency to diverge over time; a drift-normalized version of the data would test whether the advantage survives when global trends are removed.
  • A natural next step is a hybrid input that concatenates interval and ordinal representations; the paper mentions combining them as future work, and the complementary strengths seen on CCC versus SDA suggest such a hybrid could dominate both metrics.
  • The authors note that their regression objective does not reflect the ordinal nature of the labels; modeling this as a preference-learning or classification problem might amplify the ordinal advantage beyond what the LSTM regression setup shows.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper proposes a new family of ambiguity-aware emotion representations for continuous emotion recognition. Three representations are compared: an interval representation I, which models the per-time-step distribution of annotator ratings via its mean and standard deviation; an individual ordinal representation OI, which models the distribution of per-annotator trace gradients; and a group ordinal representation OG, which models the temporal gradients of the group-level distribution parameters (dµ/dt and dσ/dt). These are integrated into an LSTM-based CER system and evaluated on RECOLA (arousal, valence; bounded labels) and GameVibe (engagement; unbounded labels) using CCC and SDA. The paper claims that ordinal representations outperform interval representations on unbounded labels and on SDA for bounded traces, and that the group representation is generally superior to the individual representation.

Significance. If the central claim were established, the framework would be a useful contribution to ambiguity-aware affective computing: it explicitly connects inter-rater ambiguity with temporal dynamics, it is clearly formalized, and it is tested on one bounded and one unbounded corpus. The use of SDA as a complementary metric is appropriate for ordinal representations, and the authors are transparent about some inconsistencies in their own results. The main conceptual strength is that the paper takes seriously the idea that emotion annotations are relative and dynamic rather than absolute point values. However, the empirical comparison as presented is not like-for-like, and several summary claims go beyond what the tables show. The framework may still be valuable, but the current evidence does not support the headline conclusions without additional re-analysis on a common target.

major comments (3)
  1. [§III-A/B, Eqs. (3), (6), (7); Tables I–II] The comparisons are not on a common prediction target. I predicts the raw distribution parameters θ_n = {µ_n, σ_n}; OG predicts their time derivatives {dµ/dt, dσ/dt}; OI predicts the distribution of per-annotator gradients. The reported CCC and SDA values are computed between each model's output and its own representation target. For example, the SDA of I compares sign changes of µ̂_t with sign changes of µ_t (first derivatives), whereas the SDA of OG compares sign changes of the predicted dµ/dt with sign changes of the true dµ/dt (second derivatives of µ). Superior ordinal performance may therefore reflect target stationarity or smoothness rather than a better emotion representation. The GameVibe σ row in Table II is a clear illustration: I achieves SDA σ = 0.586 while OG achieves 0.112, consistent with raw σ drifting monotonically (Fig. 4) while dσ/dt is oscillatory. To support the abs
  2. [§V-C, §VI-A, Table I] RECOLA results are based on a single train/development split with no repeated runs, confidence intervals, or significance tests. The highlighted differences (e.g., arousal SDA µ: I=0.430, OI=0.482, OG=0.506; valence SDA µ: I=0.130, OI=0.221) may be within run-to-run or split-to-split variability. The paper should report variance over LSTM seeds and/or cross-validation folds and provide pairwise significance tests or effect sizes before ranking representations.
  3. [Abstract and §VI-B, Table II] The abstract states that ordinal representations 'outperform conventional ambiguity-aware models on unbounded labels, achieving the highest CCC and SDA scores' and 'excel in SDA' for bounded traces. Table II shows OG has the lowest SDA σ (0.112 vs I 0.586 and OI 0.534). On RECOLA, ordinal representations are better only on SDA µ, not on SDA σ, and I is substantially better on CCC. The discussion in §VI-B partially acknowledges the σ anomaly, but the abstract and the closing claims are stronger than the evidence. The claims should be revised to match the partial pattern, or a statistical basis for declaring overall superiority should be provided.
minor comments (5)
  1. [Fig. 4 caption] Typo: 'the sold line' should be 'the solid line'.
  2. [§VII] Typo: 'alternative models beyond LSMTs' should be 'beyond LSTMs'.
  3. [§VIII] Incomplete sentence: 'it is essential to for improving the generalizability and fairness' should be revised.
  4. [Author block / metadata] There is a visible encoding artifact after the author line (the '/uni00000013/...' sequence) that should be removed before publication.
  5. [§II-B] The claim that 'people agree more on temporal change than on magnitude' is central to the motivation but is supported only by citations. Consider adding a small empirical demonstration from RECOLA/GameVibe (e.g., inter-annotator agreement on sign of change vs. raw value) to make the premise more concrete.

Circularity Check

1 steps flagged · score 2.0 of 10

Empirical comparisons are self-contained; only a minor motivational self-citation in the ordinal-affect premise.

  1. other [Section II-B, Ordinal Affect Modeling]
    "Human affective assessments are inherently relative and subject to contextual factors and different anchoring and recency effects [18], [24], [25]. Therefore, ordinal models offer a more robust representation, aligning better with our own cognitive processes when we annotate affect."

    The premise that affect is ordinal/relative, which motivates defining ordinal representations as the rate of change, is partly supported by [18] (Yannakakis et al., an author of this paper) rather than by an independent, machine-checked or externally established result. This is a self-citation used in the motivation. It is not load-bearing for the reported CCC/SDA numbers, which are computed from external corpora and could have favored the interval baseline; it is also backed by external references [20], [21], [24]-[27]. Thus it is a minor self-referential element, not a construction-level circularity.

full rationale

The paper's central claims are empirical. Each representation defines a prediction target: I predicts raw distribution parameters (Eq. 3), OG predicts their time-derivatives (Eq. 7), and OI predicts the distribution of per-annotator gradients (Eqs. 4-6). The LSTM models are trained and evaluated on those respective targets using external RECOLA and GameVibe data. The higher CCC/SDA of OG/OI on their own targets is a data-dependent outcome, not an algebraic consequence of the definitions; the models could have failed to track the derivative targets. The skeptic's target-mismatch concern is a valid evaluation-design caveat, but it is not circularity because no reported number is identical by construction to a fitted parameter or to a self-cited result. The only self-referential element is the motivational claim that human affect is inherently ordinal, which leans partly on the authors' prior work [18], [19], [22], [28] alongside external sources; this does not determine the experimental outcomes. The paper also discloses limitations (regression framing not ordinal, normalization sensitivity), further supporting a low circularity score.

Assumptions & free parameters 2 free parameters · 3 assumptions · 0 invented entities

The central claim rests on the domain assumption that ordinal (relative) judgments are more reliable, plus the choice of distribution families and the window parameter F. No new physical or mathematical entities are introduced.

free parameters (2)
  • F (neighboring frame window) = 1 (RECOLA)
    Number of neighboring frames concatenated for distribution fitting in Eq. (2); set to 1 without a reported sweep, and may influence the smoothness and quality of the interval parameters from which ordinal representations are derived.
  • BoAW codebook size = 100 clusters
    Size of the bag-of-audio-words codebook for RECOLA audio features; chosen following prior work and not optimized here.
assumptions (3)
  • domain assumption Human affective assessments are inherently relative; people agree more on temporal changes than on absolute magnitudes.
    Underlies the entire ordinal representation; cited to refs [18]-[22] in Section II-B but not independently validated in this study.
  • domain assumption Annotation distributions are adequately captured by Beta (bounded traces) and Gaussian (unbounded traces) distributions.
    Required to define the interval representation I whose gradients yield OG; follows prior work [13], [38].
  • standard math The central-difference approximation (Eq. 5) provides an unbiased estimate of the temporal gradient of the traces.
    Used to compute all ordinal representations; standard but imposes smoothness assumptions on the traces.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Emotions as Ambiguity-aware Ordinal Representations." pith.science (2026). https://pith.science/paper/DBLLXLEO

@misc{pith2026250819193,
  author       = {Pith},
  title        = {Pith review of: Emotions as Ambiguity-aware Ordinal Representations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DBLLXLEO}},
  note         = {Machine review of arXiv:2508.19193}
}
read the original abstract

Emotions are inherently ambiguous and dynamic phenomena, yet existing continuous emotion recognition approaches either ignore their ambiguity or treat ambiguity as an independent and static variable over time. Motivated by this gap in the literature, in this paper we introduce ambiguity-aware ordinal emotion representations, a novel framework that captures both the ambiguity present in emotion annotation and the inherent temporal dynamics of emotional traces. Specifically, we propose approaches that model emotion ambiguity through its rate of change. We evaluate our framework on two affective corpora -- RECOLA and GameVibe -- testing our proposed approaches on both bounded (arousal, valence) and unbounded (engagement) continuous traces. Our results demonstrate that ordinal representations outperform conventional ambiguity-aware models on unbounded labels, achieving the highest Concordance Correlation Coefficient (CCC) and Signed Differential Agreement (SDA) scores, highlighting their effectiveness in modeling the traces' dynamics. For bounded traces, ordinal representations excel in SDA, revealing their superior ability to capture relative changes of annotated emotion traces.

Figures

Figures reproduced from arXiv: 2508.19193 by the authors.

Figure 1
Figure 1. High-level conceptual visualization of ambiguity-aware [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. An example of the different emotion representations as employed [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. The three different emotion representations detailed in [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (2 more)
Figure 3
Figure 3. Figure 3: A visual representation of the methodology followed. Traces from the RECOLA [5] (arousal, valence) and GameVibe [4] (engagement) corpora [PITH_FULL_IMAGE:figures/full_fig_p006_3.png]
Figure 4
Figure 4. Figure 4: An example of signal dissociation over time between annotators in [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

49 extracted references · 48 canonical work pages

  1. [1]

    A systematic review on affective computing: Emotion models, databases, and recent advances,

    Y . Wang, W. Song, W. Tao, A. Liotta, D. Yang, X. Li, S. Gao, Y . Sun, W. Ge, W. Zhang et al., “A systematic review on affective computing: Emotion models, databases, and recent advances,” Information Fusion, 2022

  2. [2]

    Basic emotions,

    P. Ekman et al., “Basic emotions,” Handbook of cognition and emotion , vol. 98, no. 45-60, p. 16, 1999

  3. [3]

    A circumplex model of affect

    J. A. Russell, “A circumplex model of affect.” Journal of personality and social psychology , vol. 39, no. 6, 1980

  4. [4]

    GameVibe: a multimodal affective game corpus,

    M. Barthet, M. Kaselimi, K. Pinitas, K. Makantasis, A. Liapis, and G. N. Yannakakis, “GameVibe: a multimodal affective game corpus,”Scientific Data, vol. 11, no. 1, p. 1306, 2024

  5. [5]

    Introducing the recola multimodal corpus of remote collaborative and affective interactions,

    F. Ringeval, A. Sonderegger, J. Sauer, and D. Lalanne, “Introducing the recola multimodal corpus of remote collaborative and affective interactions,” in 2013 10th IEEE international conference and workshops on automatic face and gesture recognition (FG) . IEEE, 2013, pp. 1–8

  6. [6]

    The Ambiguous World of Emotion Representation

    V . Sethu, E. M. Provost, J. Epps, C. Busso, N. Cummins, and S. Narayanan, “The ambiguous world of emotion representation,” arXiv preprint arXiv:1909.00360, 2019

  7. [7]

    Automatic, dimensional and continuous emo- tion recognition,

    H. Gunes and M. Pantic, “Automatic, dimensional and continuous emo- tion recognition,” International Journal of Synthetic Emotions (IJSE) , vol. 1, no. 1, pp. 68–99, 2010

  8. [8]

    Categorical and dimensional affect analysis in continuous input: Current trends and future directions,

    H. Gunes and B. Schuller, “Categorical and dimensional affect analysis in continuous input: Current trends and future directions,” Image and Vision Computing, vol. 31, no. 2, pp. 120–136, 2013

Show all 49 references
  1. [9]

    A VEC 2015: The 5th international audio/visual emotion challenge and work- shop,

    F. Ringeval, B. Schuller, M. Valstar, R. Cowie, and M. Pantic, “A VEC 2015: The 5th international audio/visual emotion challenge and work- shop,” in Proceedings of the 23rd ACM international conference on Multimedia, 2015, pp. 1335–1336

  2. [10]

    End-to-end speech emotion recognition using deep neural networks,

    P. Tzirakis, J. Zhang, and B. W. Schuller, “End-to-end speech emotion recognition using deep neural networks,” in 2018 IEEE international conference on acoustics, speech and signal processing (ICASSP). IEEE, 2018, pp. 5089–5093

  3. [11]

    From hard to soft: Towards more human-like emotion recognition by modelling the perception uncertainty,

    J. Han, Z. Zhang, M. Schmitt, M. Pantic, and B. Schuller, “From hard to soft: Towards more human-like emotion recognition by modelling the perception uncertainty,” in Proceedings of the 25th ACM international conference on Multimedia , 2017, pp. 890–897

  4. [12]

    Dynamic multi-rater gaussian mixture regression incorporating temporal dependencies of emotion uncertainty using kalman filters,

    T. Dang, V . Sethu, and E. Ambikairajah, “Dynamic multi-rater gaussian mixture regression incorporating temporal dependencies of emotion uncertainty using kalman filters,” in2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) . IEEE, 2018, pp...

  5. [13]

    Parametric distributions to model numerical emotion labels,

    D. Bose, V . Sethu, and E. Ambikairajah, “Parametric distributions to model numerical emotion labels,” Proc. Interspeech 2021 , pp. 4498– 4502, 2021

  6. [14]

    A novel sequential Monte Carlo framework for predicting ambiguous emotion states,

    J. Wu, T. Dang, V . Sethu, and E. Ambikairajah, “A novel sequential Monte Carlo framework for predicting ambiguous emotion states,” in ICASSP 2022-2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) . IEEE, 2022, pp. 8567–8571

  7. [15]

    Continuous emotion ambiguity prediction: Modeling with beta distributions,

    D. Bose, V . Sethu, and E. Ambikairajah, “Continuous emotion ambiguity prediction: Modeling with beta distributions,” IEEE Transactions on Affective Computing, 2024

  8. [16]

    Can modelling inter-rater ambiguity lead to noise-robust continuous emotion predictions?

    Y .-T. Wu, J. Wu, V . Sethu, and C.-C. Lee, “Can modelling inter-rater ambiguity lead to noise-robust continuous emotion predictions?” inProc. Interspeech 2024, 2024, pp. 3714–3718

  9. [17]

    Dual-constrained dynamical neural odes for ambiguity-aware continuous emotion predic- tion,

    J. Wu, T. Dang, V . Sethu, and E. Ambikairajah, “Dual-constrained dynamical neural odes for ambiguity-aware continuous emotion predic- tion,” arXiv preprint arXiv:2407.21344 , 2024

  10. [18]

    The ordinal nature of emotions: An emerging approach,

    G. N. Yannakakis, R. Cowie, and C. Busso, “The ordinal nature of emotions: An emerging approach,” IEEE Transactions on Affective Computing (Early Access) , 2018

  11. [19]

    The ordinal nature of emotions,

    ——, “The ordinal nature of emotions,” in Proceedings of the Intl. Conference on Affective Computing and Intelligent Interaction . IEEE, 2017, pp. 248–255

  12. [20]

    Indirect scaling methods for testing quantitative emotion theories,

    M. Junge and R. Reisenzein, “Indirect scaling methods for testing quantitative emotion theories,” Cognition & Emotion , vol. 27, no. 7, pp. 1247–1275, 2013

  13. [21]

    The relativity of ‘absolute’judgements,

    D. Laming, “The relativity of ‘absolute’judgements,” British Journal of Mathematical and Statistical Psychology , vol. 37, no. 2, pp. 152–183, 1984

  14. [22]

    From interval to ordinal: A HMM based approach for emotion label conversion,

    J. Wu, T. Dang, V . Sethu, and E. Ambikairajah, “From interval to ordinal: A HMM based approach for emotion label conversion,” in Proc. Interspeech 2023, 2023, pp. 1843–1847

  15. [23]

    The impact of rewards and trait reward responsiveness on player motivation,

    C. Phillips, D. Johnson, M. Klarkowski, M. J. White, and L. Hides, “The impact of rewards and trait reward responsiveness on player motivation,” in Proceedings of the 2018 Symposium on Computer-Human Interaction in Play. ACM, 2018, pp. 393–404

  16. [24]

    A. R. Damasio, Descartes’ error: Emotion, rationality and the human brain. New York: Putnam, 1994

  17. [25]

    Anchors, scales and the relative coding of value in the brain,

    B. Seymour and S. M. McClure, “Anchors, scales and the relative coding of value in the brain,” Current opinion in neurobiology , vol. 18, no. 2, pp. 173–178, 2008

  18. [26]

    The relation between culture and response styles: Evidence from 19 countries,

    T. Johnson, P. Kulesa, Y . I. Cho, and S. Shavitt, “The relation between culture and response styles: Evidence from 19 countries,” Journal of Cross-cultural psychology, vol. 36, no. 2, pp. 264–277, 2005

  19. [27]

    The framing of decisions and the psychology of choice,

    A. Tversky and D. Kahneman, “The framing of decisions and the psychology of choice,” Science, vol. 211, no. 4481, pp. 453–458, 1981

  20. [28]

    A study on affect model validity: Nominal vs ordinal labels,

    D. Melhart, K. Sfikas, G. Giannakakis, and G. Y . A. Liapis, “A study on affect model validity: Nominal vs ordinal labels,” in Workshop on Artificial Intelligence in Affective Computing . Proceedings of Machine Learning Research, 2020, pp. 27–34

  21. [29]

    Dynamic difficulty awareness training for continuous emotion prediction,

    Z. Zhang, J. Han, E. Coutinho, and B. Schuller, “Dynamic difficulty awareness training for continuous emotion prediction,” IEEE Transac- tions on Multimedia , vol. 21, no. 5, pp. 1289–1301, 2018

  22. [30]

    An investigation of annotation delay compensation and output- associative fusion for multimodal continuous emotion prediction,

    Z. Huang, T. Dang, N. Cummins, B. Stasak, P. Le, V . Sethu, and J. Epps, “An investigation of annotation delay compensation and output- associative fusion for multimodal continuous emotion prediction,” in Proceedings of the 5th International Workshop on Audio/Visual Emotion Ch...

  23. [31]

    Ranktrace: Relative and unbounded affect annotation,

    P. Lopes, G. N. Yannakakis, and A. Liapis, “Ranktrace: Relative and unbounded affect annotation,” in Proceedings of the Intl. Conference on Affective Computing and Intelligent Interaction . IEEE, 2017, pp. 158–163

  24. [32]

    PAGAN: Video affect annotation made easy,

    D. Melhart, A. Liapis, and G. N. Yannakakis, “PAGAN: Video affect annotation made easy,” in Proceedings of the Conference on Affective Computing and Intelligent Interaction (ACII) , 2019

  25. [33]

    Varying the context to advance affect modelling: A study on game engagement prediction,

    K. Pinitas, N. Rasajski, M. Barthet, M. Kaselimi, K. Makantasis, A. Liapis, and G. N. Yannakakis, “Varying the context to advance affect modelling: A study on game engagement prediction,” in Proceedings of the Conference on Affective Computing and Intelligent Interaction (ACII), 2024

  26. [34]

    Across- game engagement modelling via few-shot learning,

    K. Pinitas, K. Makantasis, and G. N. Yannakakis, “Across- game engagement modelling via few-shot learning,” arXiv preprint arXiv:2409.13002, 2024

  27. [35]

    At the border of acoustics and linguistics: Bag-of-audio-words for the recognition of emotions in speech

    M. Schmitt, F. Ringeval, and B. W. Schuller, “At the border of acoustics and linguistics: Bag-of-audio-words for the recognition of emotions in speech.” in Interspeech, 2016, pp. 495–499

  28. [36]

    openxbow–introducing the passau open- source crossmodal bag-of-words toolkit,

    M. Schmitt and B. Schuller, “openxbow–introducing the passau open- source crossmodal bag-of-words toolkit,” Journal of Machine Learning Research, vol. 18, no. 96, pp. 1–5, 2017

  29. [37]

    Videomae v2: Scaling video masked autoencoders with dual masking,

    L. Wang, B. Huang, Z. Zhao, Z. Tong, Y . He, Y . Wang, Y . Wang, and Y . Qiao, “Videomae v2: Scaling video masked autoencoders with dual masking,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2023, pp. 14 549–14 560

  30. [38]

    Predicting the distribution of emotion perception: capturing inter-rater variability,

    B. Zhang, G. Essl, and E. Mower Provost, “Predicting the distribution of emotion perception: capturing inter-rater variability,” in Proceedings of the 19th ACM International Conference on Multimodal Interaction , 2017, pp. 51–59

  31. [39]

    Using gaussian processes with lstm neural networks to predict continuous-time, dimensional emotion in ambiguous speech,

    M. Atcheson, V . Sethu, and J. Epps, “Using gaussian processes with lstm neural networks to predict continuous-time, dimensional emotion in ambiguous speech,” in 2019 8th International Conference on Affective Computing and Intelligent Interaction (ACII) . IEEE, 2019, pp. 718– 724

  32. [40]

    A concordance correlation coefficient to evaluate reproducibility,

    I. Lawrence and K. Lin, “A concordance correlation coefficient to evaluate reproducibility,” Biometrics, pp. 255–268, 1989

  33. [41]

    Fifty shades of green: Towards a robust measure of inter-annotator agreement for continuous signals,

    B. M. Booth and S. S. Narayanan, “Fifty shades of green: Towards a robust measure of inter-annotator agreement for continuous signals,” in Proceedings of the 2020 international conference on multimodal interaction, 2020, pp. 204–212

  34. [42]

    Multimodal affect models: An investigation of relative salience of audio and visual cues for emotion prediction,

    J. Wu, T. Dang, V . Sethu, and E. Ambikairajah, “Multimodal affect models: An investigation of relative salience of audio and visual cues for emotion prediction,” Frontiers in Computer Science , vol. 3, p. 767767, 2021

  35. [43]

    Real-world automatic continu- ous affect recognition from audiovisual signals,

    P. Tzirakis, S. Zafeiriou, and B. Schuller, “Real-world automatic continu- ous affect recognition from audiovisual signals,” in Multimodal behavior analysis in the wild . Elsevier, 2019, pp. 387–406

  36. [44]

    V ocal expression and perception of emotion,

    J.-A. Bachorowski, “V ocal expression and perception of emotion,” Current directions in psychological science , vol. 8, no. 2, pp. 53–57, 1999

  37. [45]

    Estimation of continuous valence and arousal levels from faces in naturalistic conditions,

    A. Toisoul, J. Kossaifi, A. Bulat, G. Tzimiropoulos, and M. Pantic, “Estimation of continuous valence and arousal levels from faces in naturalistic conditions,” Nature Machine Intelligence , vol. 3, no. 1, pp. 42–50, 2021

  38. [46]

    Abaw: Valence-arousal estimation, expression recognition, action unit detection & emotional reaction intensity estimation challenges,

    D. Kollias, P. Tzirakis, A. Baird, A. Cowen, and S. Zafeiriou, “Abaw: Valence-arousal estimation, expression recognition, action unit detection & emotional reaction intensity estimation challenges,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recog...

  39. [47]

    Moment-to-moment engagement prediction through the eyes of the observer: Pubg streaming on twitch,

    D. Melhart, D. Gravina, and G. N. Yannakakis, “Moment-to-moment engagement prediction through the eyes of the observer: Pubg streaming on twitch,” in Proceedings of the 15th International Conference on the Foundations of Digital Games , 2020, pp. 1–10

  40. [48]

    Towards general models of player experience: A study within genres,

    D. Melhart, A. Liapis, and G. N. Yannakakis, “Towards general models of player experience: A study within genres,” in 2021 IEEE Conference on Games (CoG) . IEEE, 2021, pp. 01–08

  41. [49]

    Towards general models of player affect,

    E. Camilleri, G. N. Yannakakis, and A. Liapis, “Towards general models of player affect,” in Proceedings of the Intl. Conference on Affective Computing and Intelligent Interaction . IEEE, 2017, pp. 333–339

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.