Pith. sign in

REVIEW 4 major objections 5 minor 54 references

Auditing Facial Emotion Recognition Datasets for Posed Expressions and Racial Bias

T0 review · 4 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read Auditing AffectNet and RAF-DB, this paper finds that a substantial share of supposedly in-the-wild images are posed, and that models trained on these datasets are more likely to predict negative emotions for smiling faces observed as…

desk verdict The racial-bias claim doesn't survive contact with the missing base-rate control, and the posed-image percentages come from a stratified sample that isn't weighted back; the question is good, but the evidence as presented doesn't support the headline. read the letter →

arxiv 2507.10755 v1 pith:7N4VL5EP submitted 2025-07-14 cs.CV cs.AI

Rina Khan , Catherine Stinson This is my paper · ORCID
classification cs.CVcs.AI
keywords facialexpressionrecognitionalgorithmicauditracialbiasposedexpressionsspontaneousAffectNetRAF-DBFairFace
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

This paper tries to establish that two widely used facial expression recognition (FER) datasets, AffectNet and RAF-DB, contain a substantial number of posed images even though they are presented as capturing spontaneous 'in-the-wild' expressions, and that models trained on them are biased against people perceived as non-white or as having darker skin. Using a proposed method for distinguishing posed from spontaneous smiles, the audit finds 46.5% of sampled AffectNet images and 35.3% of sampled RAF-DB images are posed. Running two top-performing models on the FairFace test set, the paper reports that faces observed as Black, East Asian, Southeast Asian, or Indian are significantly more likely than White faces to be predicted as angry, sad, disgusted, or contemptuous while smiling or neutral, with the effect increasing for darker skin tones. If true, these findings matter because real-world uses of emotion recognition—automated interviews, security, and hiring—could both lose accuracy on genuine expressions and perpetuate harmful misreadings of people's emotions based on appearance.

What carries the argument

The central object is the smile classifier based on the Facial Action Coding System (FACS): a genuine smile is indicated by Action Units 6 (cheek raising), 12 (lip corner pull), and 25 (parted lips), while a posed smile lacks AU6. For non-smiling images, the audit treats recognizable actors, plain monochrome backgrounds, and bright artificial lighting combined with direct camera gaze as markers of posing. On the fairness side, the load-bearing metric is the proportion of smiling or neutral faces inside the set of negative emotion predictions, stratified by observed race from the FairFace test set and by a six-point skin-tone scale grouped into three bins, with White faces serving as the reference category for chi-square tests. This metric carries the bias conclusion: if models were unbiased, the smiling rate among negative predictions should be the same across observed races.

What would settle it

Re-run the bias audit on a subset of test images where the base rate of smiling and neutral faces is matched across observed-race groups, or where smile/neutral labels are independently verified by multiple annotators; if the gap in negative predictions on smiling faces between White and non-White groups disappears or reverses, the claimed bias pattern would be shown to reflect test-set demographics rather than systematic model bias.

Watch

Extended reading notes

Core claim

The paper's core assertion is that the audited FER systems are racially biased and their training data are not genuinely spontaneous. On the data side, the authors propose a FACS-inspired rule for smiles (presence of AU6/AU12/AU25 for genuine, absence of AU6 for posed) plus background and gaze heuristics for non-smiles, and apply it to random samples of AffectNet and RAF-DB, concluding that 46.5% of AffectNet and 35.3% of RAF-DB images are posed, a result statistically significant against a null hypothesis of zero posed images. On the model side, the paper runs the Relative Uncertainty Learning model and the Multi-task EfficientNet-B2 model on FairFace, then compares the rate of negative emotion predictions (anger, sadness, disgust, contempt) among faces the auditors label as smiling or neutral. Across both models, 23.4% of negative predictions on White faces are smiling, versus 33% on Black faces and higher on several other non-White groups, with significant chi-square differences for Black, East Asian, Southeast Asian, and Indian observed races; the neutral-face and skin-tone analyses show a similar or stronger pattern. The paper interprets these gaps as evidence that the models over-assign negative emotions to non-White and darker-skinned faces, arguing that a smiling or neutral face should not receive a negative label.

Load-bearing premise

The bias claim hinges on treating a smiling or neutral face as a ground-truth label that cannot legitimately receive a negative emotion prediction, and on trusting the auditors' manual smile/neutral labels; if smiling simply occurs at different frequencies across observed-race groups in the test set, the reported gaps could be an artifact of those base-rate differences rather than model bias.

Editorial extensions

If this is right

  • Benchmark accuracy of these models on 'in-the-wild' data overstates real-world performance, because a large portion of their training images are posed rather than spontaneous expressions.
  • Deployed emotion-recognition systems risk misclassifying smiling or neutral faces of non-White and darker-skinned people as angry, sad, disgusted, or contemptuous, affecting automated interviews, security, and hiring.
  • Underrepresentation of dark skin tones in the training samples (about 4% in the darkest group) likely contributes to the bias, though the sample is too small to test annotation bias.
  • The proposed posed-image detection methodology can be reused by other auditors to check spontaneity in FER image datasets.

Reading between the lines

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

  • A stratified re-analysis that controls for how often each observed-race group smiles in FairFace could separate true model bias from base-rate effects, something the paper does not report.
  • Because a large share of training images are posed, models trained on these datasets may be tuned to exaggerated, actor-like expressions; evaluating on a corpus of genuinely spontaneous interactions would likely show a larger accuracy drop for non-White faces.
  • The posed-image detection method, built for smiles, could be extended to video FER datasets and to non-smile emotions, where FACS markers are less established, to test whether posing affects other expressions similarly.
  • A stronger causal test of skin-tone sensitivity would manipulate perceived skin tone in images while holding expression and identity fixed, isolating the model's reliance on skin cues from the demographic composition of training data.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The manuscript audits two facial expression recognition (FER) datasets, AffectNet and RAF-DB, and two models trained on them (RUL and MENet). It makes two central claims: first, that a large fraction of supposedly in-the-wild images in these datasets are actually posed (46.5% in AffectNet and 35.3% in RAF-DB, based on a custom visual heuristic); second, that the audited FER models are racially biased because, when the models predict a negative emotion, the proportion of smiling or neutral faces is higher for people observed as non-white or with darker skin than for people observed as white. The paper proposes this as the first audit to verify the presence of posed expressions in FER datasets and to examine fairness across more than two observed-race groups.

Significance. If the claims were established, this would be a valuable contribution to the algorithmic-auditing literature: the topic is timely, the use of FairFace with multiple observed-race categories is a step beyond binary race audits, and the paper makes falsifiable, quantitative claims rather than only qualitative arguments. The paper also usefully foregrounds the connection between data-collection practices and downstream model behavior. However, as presented, the evidence does not support the central claims. The bias analysis compares a reverse conditional—P(smiling | negative prediction, race)—without controlling for how often each group smiles in the evaluation set, so a race-blind model could produce the same pattern solely from group differences in smile base rates. The smile/neutral labels used as ground truth are unvalidated and appear to be assigned without inter-rater reliability or blinding. The posed-expression heuristic is similarly unvalidated and partly circular, and the reported statistical intervals are inconsistent with the stratified sampling design.

major comments (4)
  1. [Section IV-B, Tables III–IV, Figures 2–3] The bias analysis compares P(smiling | negative prediction, observed race) across groups and treats the white group's value as the no-bias null. This is valid only if smile prevalence in FairFace is equal across observed-race groups, which the paper does not establish and which is unlikely to hold in a naturalistic dataset. A race-blind classifier whose errors are independent of race would still produce the reported pattern if, for example, Black or Southeast Asian faces are more often smiling in FairFace, because more smiling faces would fall into any negative-prediction bin. The authors should report P(smile | race) on the full evaluation set and the conditional error rates P(negative prediction | smile, race) and P(negative prediction | non-smile, race) with confidence intervals; the reverse conditional alone cannot support the bias conclusion. The same concern applies to the neutral-face analysis and to the skin-tone analysis in Figure 3. Additionally, because inference outputs were sampled by predicted emotion and observed race, the reported marginal proportions are not representative of the models' behavior on FairFace as a whole.
  2. [Section III-F] The smile and neutral labels used as ground truth for the bias claim are assigned by the authors with no inter-rater reliability, no FACS verification, and no reported blinding to model predictions or observed race. Because the audit hypothesis is known, this creates a risk of systematic mislabeling that could generate the reported differences. The paper also assumes, without support, that a smiling or neutral face cannot validly receive a negative emotion prediction; this ignores display rules, social context, and the possibility of genuine smiles accompanying negative affect. At minimum, the authors need an independent label validation (for example, multiple coders with agreement statistics) and a sensitivity analysis using only high-confidence labels.
  3. [Section III-B and Section IV-A] The posed-expression methodology is not validated. The FACS-based smile criterion applies only to smiles; for non-smiling faces, the paper relies on visual heuristics (recognizable actors, plain monochrome backgrounds, direct gaze, artificial lighting) that are asserted without evidence or annotation reliability. Because these same heuristics partly define what the auditors count as posed, the claim that 46.5% of AffectNet and 35.3% of RAF-DB images are posed is to a large extent a restatement of the labeling rule rather than an empirical discovery. The null hypothesis of "zero posed images in a completely wild setting" is also not an appropriate benchmark for in-the-wild datasets, and the chi-square test against it does not validate the heuristic. The authors should validate their protocol on a dataset with known spontaneous and posed labels and report per-criterion precision and recall.
  4. [Section III-A and Figure 1] The statistical reporting for the posed-image proportions is inconsistent with the sampling design. The sample size is initially computed under simple random sampling with a 95% confidence level and 5% margin of error, but the authors then state that they "sample equal numbers of each emotion label," which is a stratified design. The confidence intervals and the combined chi-square test with N = 761 ignore this stratification and are therefore not valid as reported. Additionally, the caption of Figure 1 gives the combined proportion as 40.9% with a 95% CI of 41.3%–51.6%, which is impossible because the point estimate lies outside the interval; the interval appears to have been copied from the AffectNet row. The authors should compute weighted estimates and report strata-specific results.
minor comments (5)
  1. [Throughout] The manuscript contains frequent typos, including "interations" (Introduction), "exppressions" (Section II-B), "Relative Uncertainity" (Table I and Section IV-B), "Southest Asian" (Table II), and "osbserved" (Table II caption).
  2. [Section III-E and Table II] The paper says it aimed to select 50 samples for each permutation of observed race and predicted emotion (Section III-E), but the row totals in Table II are not multiples of 50 for some groups; please clarify the actual sampling and inclusion criteria.
  3. [Figures 2 and 3] The bar charts report proportions without error bars or confidence intervals, so the reader cannot assess the precision of the between-group differences; adding these would help interpretation.
  4. [Tables III and IV] The multiple chi-square tests are not corrected for multiple comparisons; with 12 tests, some significant results would be expected by chance even under the null hypothesis.
  5. [Section V] The statement that "only 4% of the sampled images were of the darkest skin tone" is not tied to any reported count or table; please provide exact numbers and clarify which sample it refers to.

Circularity Check

1 steps flagged · score 2.0 of 10

Posed-expression audit is mildly self-definitional, but the central bias analysis is empirical and not circular.

  1. self definitional [Section III-B (Spontaneous versus Posed) and Section IV-A (Spontaneous versus Posed)]
    "We also consider the background of the image. It is common for photographers to pose models on plain monocolor backgrounds to use their images as stock images. ... by identifying cases of plain mono-color backgrounds we can identify images where we have a high confidence of the image being posed. ... Finally we identify if a person is looking straight at the camera and if they are very well lit without being out in the sunlight. ... When analyzing samples from datasets, we found a large number of images to be posed."

    The audit's operational rule for classifying a non-smiling image as posed is the presence of the very cues that are then counted: recognizable actors, plain monocolor backgrounds, direct camera gaze, and artificial lighting. Consequently, the Section IV-A result that 46.5% of AffectNet and 35.3% of RAF-DB images are posed is not an independent empirical discovery but a restatement of the rubric applied to the samples. Under that rubric, an image with a plain background and direct gaze is by definition high-confidence posed, so the chi-square test against a 'zero posed in the wild' null hypothesis tests whether the rubric fires, not whether the datasets contain acted or stock material in a theory-independent sense.

full rationale

Aside from the posed-expression operationalization, the paper's central bias claim in Section IV-B is an empirical audit rather than a derivation: the authors run FER models over FairFace, label smiles and neutral expressions, and compare the proportion of smiling faces among negative emotion predictions across observed race and skin tone. No parameter is fitted to the conclusion, and the paper does not rely on self-citations or imported uniqueness theorems. The base-rate and label-validation concerns raised by a skeptical reader are validity threats: the null hypothesis that P(smiling | negative prediction) should equal the White rate assumes race-balanced smile prevalence in FairFace and accurate smile labels, but these concerns do not make the result circular because the reported numbers could be confounded or wrong without being constructed from the conclusion. Therefore the only circular element is the mild self-definitional step in the posed detection rubric, which warrants a low score of 2 rather than 0.

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

The central claims rest on unvalidated manual annotations and uncontrolled statistical comparisons. The ledger records the domain assumptions in the posed/spontaneous and bias analyses; no parameters are fitted.

assumptions (4)
  • domain assumption A smiling or neutral face is a correct label that should not be predicted as a negative emotion.
    Used to define bias in Section IV-B; the paper counts negative predictions on smiling faces as errors without allowing for smiles in contempt, disgust, or masking, and without an independent ground truth.
  • ad hoc to paper The visual heuristics (plain background, recognizable actor, direct gaze, artificial lighting) reliably indicate posed expressions.
    Introduced in Section III-B; these cues are not validated against a ground-truth dataset and no inter-rater reliability is reported, so the posed proportion is largely an artifact of the heuristic definition.
  • domain assumption AU6, AU12 and AU25 can be accurately identified by eye from still images.
    Section III-B and III-F; FACS coding normally requires training and multiple coders, particularly on compressed web images.
  • domain assumption Fitzpatrick skin type can be reliably observed from single photographs.
    Sections III-C and III-G acknowledge lighting variation; no reliability check is reported, so the skin-tone groups may be noisy.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Auditing Facial Emotion Recognition Datasets for Posed Expressions and Racial Bias." pith.science (2026). https://pith.science/paper/7N4VL5EP

@misc{pith2026250710755,
  author       = {Pith},
  title        = {Pith review of: Auditing Facial Emotion Recognition Datasets for Posed Expressions and Racial Bias},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7N4VL5EP}},
  note         = {Machine review of arXiv:2507.10755}
}
read the original abstract

Facial expression recognition (FER) algorithms classify facial expressions into emotions such as happy, sad, or angry. An evaluative challenge facing FER algorithms is the fall in performance when detecting spontaneous expressions compared to posed expressions. An ethical (and evaluative) challenge facing FER algorithms is that they tend to perform poorly for people of some races and skin colors. These challenges are linked to the data collection practices employed in the creation of FER datasets. In this study, we audit two state-of-the-art FER datasets. We take random samples from each dataset and examine whether images are spontaneous or posed. In doing so, we propose a methodology for identifying spontaneous or posed images. We discover a significant number of images that were posed in the datasets purporting to consist of in-the-wild images. Since performance of FER models vary between spontaneous and posed images, the performance of models trained on these datasets will not represent the true performance if such models were to be deployed in in-the-wild applications. We also observe the skin color of individuals in the samples, and test three models trained on each of the datasets to predict facial expressions of people from various races and skin tones. We find that the FER models audited were more likely to predict people labeled as not white or determined to have dark skin as showing a negative emotion such as anger or sadness even when they were smiling. This bias makes such models prone to perpetuate harm in real life applications.

Figures

Figures reproduced from arXiv: 2507.10755 by the authors.

Figure 1
Figure 1. Percentage of posed images in datasets with 95% confidence interval. [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 3
Figure 3. Percentage of negative emotion predictions in samples that are [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

54 extracted references · 50 canonical work pages

  1. [1]

    B., N ELSON , A

    A DAMS , R. B., N ELSON , A. J., S OTO, J. A., H ESS , U., AND KLECK , R. E. Emotion in the Neutral Face: A Mechanism for Impression Formation? Cognition & emotion 26 , 3 (2012), 431–441

    work page 2012

  2. [2]

    F., A DOLPHS , R., M ARSELLA , S., M ARTINEZ , A

    B ARRETT , L. F., A DOLPHS , R., M ARSELLA , S., M ARTINEZ , A. M., AND POLLAK , S. D. Emotional Expressions Reconsidered: Challenges to Inferring Emotion From Human Facial Movements. Psychological Science in the Public Interest: A Journal of the American Psychological Society 20 , 1 (July 2019), 1–68

    work page 2019

  3. [3]

    S., L ITTLEWORT , G

    B ARTLETT , M. S., L ITTLEWORT , G. C., F RANK , M. G., L AINSCSEK , C., F ASEL , I. R., AND MOVELLAN , J. R. Automatic Recognition of Facial Actions in Spontaneous Expressions. JOURNAL OF MULTI- MEDIA 1 , 6 (2006)

    work page 2006

  4. [4]

    F., S RINIVASAN , R., AND MARTINEZ , A

    B ENITEZ -QUIROZ , C. F., S RINIVASAN , R., AND MARTINEZ , A. M. EmotioNet: An Accurate, Real-Time Algorithm for the Automatic Annotation of a Million Facial Expressions in the Wild. In 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (June 2016), pp. 5562–5570. ISSN: 1063-6919

    work page 2016

  5. [5]

    Proximity Begins with a Smile, But Which One? Associating Non-duchenne Smiles with Higher Psy- chological Distance

    B OGODISTOV , Y., AND DOST, F. Proximity Begins with a Smile, But Which One? Associating Non-duchenne Smiles with Higher Psy- chological Distance. Frontiers in Psychology 8 (Aug. 2017), 272863. Publisher: Frontiers

    work page 2017

  6. [6]

    Performing blackness, forming whiteness: Linguistic minstrelsy in Hollywood film1

    B UCHOLTZ , M., AND LOPEZ , Q. Performing blackness, forming whiteness: Linguistic minstrelsy in Hollywood film1. Journal of Sociolinguistics 15 , 5 (2011), 680–706. eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1111/j.1467- 9841.2011.00513.x

    arXiv 2011

  7. [7]

    Gender Shades: Intersectional Accuracy Disparities in Commercial Gender Classification

    B UOLAMWINI , J., AND GEBRU , T. Gender Shades: Intersectional Accuracy Disparities in Commercial Gender Classification. In Confer- ence on Fairness, Accountability and Transparency (Jan. 2018), PMLR, pp. 77–91. ISSN: 2640-3498

    work page 2018

  8. [8]

    Are smiles a sign of happiness? Spontaneous expressions of judo winners

    C RIVELLI , C., C ARRERA , P., AND FERN ´ANDEZ -D OLS , J.-M. Are smiles a sign of happiness? Spontaneous expressions of judo winners. Evolution and Human Behavior 36 , 1 (Jan. 2015), 52–58

    work page 2015

Show all 54 references
  1. [9]

    Imagenet: A large-scale hierarchical image database

    D ENG , J., D ONG , W., S OCHER , R., L I, L.-J., L I, K., AND FEI-F EI, L. Imagenet: A large-scale hierarchical image database. In 2009 IEEE Conference on Computer Vision and Pattern Recognition (2009), pp. 248–255

    2009

  2. [10]

    Acted Facial Expressions In The Wild Database

    D HALL , A., G OECKE , R., L UCEY, S., AND GEDEON , T. Acted Facial Expressions In The Wild Database

  3. [11]

    A., AND GEVERS , T

    D IBEKLIO ˘GLU , H., S ALAH , A. A., AND GEVERS , T. Recognition of genuine smiles. IEEE Transactions on Multimedia 17 , 3 (2015), 279– 294

    2015

  4. [12]

    E KMAN , P., AND FRIESEN , W. V. Constants across cultures in the face and emotion. Journal of personality and social psychology 17 , 2 (1971), 124

    1971

  5. [13]

    E KMAN , P., AND FRIESEN , W. V. Facial Action Coding System

  6. [14]

    F ITZPATRICK , T. B. The validity and practicality of sun-reactive skin types i through vi. Archives of dermatology 124 , 6 (1988), 869–871

    1988

  7. [15]

    G., E KMAN , P., AND FRIESEN , W

    F RANK , M. G., E KMAN , P., AND FRIESEN , W. V. Behavioral markers and recognizability of the smile of enjoyment. Journal of Personality and Social Psychology 64 , 1 (1993), 83–93. Place: US Publisher: American Psychological Association

    1993

  8. [16]

    F RIDLUND , A. J. Human facial expression: An evolutionary view. Human facial expression: An evolutionary view. Academic Press, San Diego, CA, US, 1994. Pages: xiv, 369

    1994

  9. [17]

    J., E RHAN , D., C ARRIER , P

    G OODFELLOW , I. J., E RHAN , D., C ARRIER , P. L., C OURVILLE , A., MIRZA , M., H AMNER , B., C UKIERSKI , W., T ANG , Y., T HALER , D., LEE, D.-H., Z HOU , Y., R AMAIAH , C., F ENG , F., L I, R., W ANG , X., ATHANASAKIS , D., S HAWE -TAYLOR , J., M ILAKOV, M., P ARK , J., IO...

    2013 arXiv

  10. [18]

    G., C OOKE , A

    H ALBERSTADT , A. G., C OOKE , A. N., G ARNER , P. W., H UGHES , S. A., O ERTWIG , D., AND NEUPERT , S. D. Racialized emotion recognition accuracy and anger bias of children’s faces. Emotion (Washington, D.C.) 22 , 3 (Apr. 2022), 403–417

    2022

  11. [19]

    Towards a Critical Race Methodology in Algorithmic Fairness

    H ANNA , A., D ENTON , E., S MART, A., AND SMITH -L OUD , J. Towards a Critical Race Methodology in Algorithmic Fairness. Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency (Jan. 2020), 501–512. arXiv: 1912.03593

    2020 arXiv

  12. [20]

    J., AND LI, X

    J IA, S., W ANG , S., H U, C., W EBSTER , P. J., AND LI, X. Detection of Genuine and Posed Facial Expressions of Emotion: Databases and Methods. Frontiers in Psychology 11 (Jan. 2021), 580287

    2021

  13. [21]

    DFEW: A Large-Scale Database for Recognizing Dynamic Facial Expressions in the Wild

    J IANG , X., Z ONG , Y., Z HENG , W., T ANG , C., X IA, W., L U, C., AND LIU, J. DFEW: A Large-Scale Database for Recognizing Dynamic Facial Expressions in the Wild. In Proceedings of the 28th ACM International Conference on Multimedia (New York, NY , USA, Oct. 2020), MM ’20, ...

    2020

  14. [22]

    Gender slopes: Counterfactual fairness for computer vision models by attribute manipulation

    J OO, J., AND K ¨ARKK ¨AINEN , K. Gender slopes: Counterfactual fairness for computer vision models by attribute manipulation. In Proceedings of the 2nd international workshop on fairness, accountability, trans- parency and ethics in multimedia (2020), pp. 1–5

    2020

  15. [23]

    Effect of Acting Experience on Emotion Expression and Recognition in V oice: Non-Actors Provide Better Stimuli than Expected

    J ¨URGENS , R., G RASS , A., D ROLET , M., AND FISCHER , J. Effect of Acting Experience on Emotion Expression and Recognition in V oice: Non-Actors Provide Better Stimuli than Expected. Journal of Nonverbal Behavior 39 , 3 (2015), 195–214

    2015

  16. [24]

    F., AND TIAN , Y

    K ANADE , T., C OHN , J. F., AND TIAN , Y. Comprehensive database for facial expression analysis. In Proceedings fourth IEEE interna- tional conference on automatic face and gesture recognition (cat. No. PR00580) (2000), IEEE, pp. 46–53

    2000

  17. [25]

    FairFace: Face Attribute Dataset for Balanced Race, Gender, and Age for Bias Measurement and Mitigation

    K ARKKAINEN , K., AND JOO, J. FairFace: Face Attribute Dataset for Balanced Race, Gender, and Age for Bias Measurement and Mitigation. In 2021 IEEE Winter Conference on Applications of Computer Vision (WACV) (Waikoloa, HI, USA, Jan. 2021), IEEE, pp. 1547–1557

    2021

  18. [26]

    K IRITCHENKO , S., AND MOHAMMAD , S. M. Examining Gender and Race Bias in Two Hundred Sentiment Analysis Systems, May 2018. arXiv:1805.04508 [cs]

    2018 arXiv

  19. [27]

    K O\LAKOWSKA , A., L ANDOWSKA , A., S ZWOCH , M., S ZWOCH , W., AND WR ´OBEL , M. R. Emotion Recognition and Its Applications. In Human-Computer Systems Interaction: Backgrounds and Applications 3, Z. S. Hippe, J. L. Kulikowski, T. Mroczek, and J. Wtorek, Eds. Springer Interna...

    2014

  20. [28]

    A., P APAIOANNOU , A., ZHAO, G., S CHULLER , B., K OTSIA , I., AND ZAFEIRIOU , S

    K OLLIAS , D., T ZIRAKIS , P., N ICOLAOU , M. A., P APAIOANNOU , A., ZHAO, G., S CHULLER , B., K OTSIA , I., AND ZAFEIRIOU , S. Deep affect prediction in-the-wild: Aff-wild database and challenge, deep architectures, and beyond. International Journal of Computer Vision 127, 6-...

    2019

  21. [29]

    A., O FODILE , I., E SCALERA , S., BARO, X., H YNIEWSKA , S., A LLIK , J., AND ANBARJAFARI , G

    K ULKARNI , K., C ORNEANU , C. A., O FODILE , I., E SCALERA , S., BARO, X., H YNIEWSKA , S., A LLIK , J., AND ANBARJAFARI , G. Automatic Recognition of Facial Displays of Unfelt Emotions, Jan

  22. [30]

    Reliable Crowdsourcing and Deep Locality-Preserving Learning for Expression Recognition in the Wild

    L I, S., D ENG , W., AND DU, J. Reliable Crowdsourcing and Deep Locality-Preserving Learning for Expression Recognition in the Wild. In 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (Honolulu, HI, July 2017), IEEE, pp. 2584–2593

    2017

  23. [31]

    F., K ANADE , T., S ARAGIH , J., A MBADAR , Z., AND MATTHEWS , I

    L UCEY, P., C OHN , J. F., K ANADE , T., S ARAGIH , J., A MBADAR , Z., AND MATTHEWS , I. The extended cohn-kanade dataset (ck+): A complete dataset for action unit and emotion-specified expression. In 2010 ieee computer society conference on computer vision and pattern recogni...

    2010

  24. [32]

    J., K AMACHI , M., AND GYOBA , J

    L YONS , M. J., K AMACHI , M., AND GYOBA , J. Coding facial ex- pressions with gabor wavelets (ivc special issue). arXiv preprint arXiv:2009.05938 (2020)

    2020 arXiv

  25. [33]

    Charac- terizing Bias in Classifiers using Generative Models, May 2019

    M CDUFF, D., M A, S., S ONG , Y., AND KAPOOR , A. Charac- terizing Bias in Classifiers using Generative Models, May 2019. arXiv:1906.11891 [cs]

    2019 arXiv

  26. [34]

    M OLLAHOSSEINI , A., H ASANI , B., AND MAHOOR , M. H. AffectNet: A Database for Facial Expression, Valence, and Arousal Computing in the Wild. IEEE Transactions on Affective Computing 10 , 1 (Jan. 2019), 18–31. arXiv:1708.03985 [cs]

    2019 arXiv

  27. [35]

    N OBLE , S. U. Algorithms of oppression. In Algorithms of oppression . New York university press, 2018

    2018

  28. [36]

    Weapons of math destruction: How big data increases inequality and threatens democracy

    O’ NEIL , C. Weapons of math destruction: How big data increases inequality and threatens democracy . Crown, 2017

    2017

  29. [37]

    Papers with Code - The latest in Machine Learning, 2022

    P APERS WITH CODE. Papers with Code - The latest in Machine Learning, 2022

    2022

  30. [38]

    The mahnob laughter database

    P ETRIDIS , S., M ARTINEZ , B., AND PANTIC , M. The mahnob laughter database. Image and Vision Computing 31 , 2 (2013), 186–202

    2013

  31. [39]

    Racial Influence on Automated Perceptions of Emotions

    R HUE , L. Racial Influence on Automated Perceptions of Emotions. SSRN Scholarly Paper ID 3281765, Social Science Research Network, Rochester, NY , Nov. 2018

    2018

  32. [40]

    R OTH, W. D. The multiple dimensions of race. Ethnic and Racial Studies 39 , 8 (June 2016), 1310–1338

    2016

  33. [41]

    R USSELL , J. A. A circumplex model of affect. Journal of Personality and Social Psychology 39 , 6 (1980), 1161–1178. Place: US Publisher: American Psychological Association

    1980

  34. [42]

    V., S AVCHENKO , L

    S AVCHENKO , A. V., S AVCHENKO , L. V., AND MAKAROV, I. Classify- ing emotions and engagement in online learning based on a single facial expression recognition neural network. IEEE Transactions on Affective Computing (2022), 1–12. Conference Name: IEEE Transactions on Affecti...

    2022

  35. [43]

    L., A MBADAR , Z., C OHN , J

    S CHMIDT , K. L., A MBADAR , Z., C OHN , J. F., AND REED , L. I. Movement Differences Between Deliberate and Spontaneous Facial Ex- pressions: Zygomaticus Major Action in Smiling. Journal of Nonverbal Behavior 30 , 1 (2006), 37–52

    2006

  36. [44]

    No Classification without Representation: Assessing Geodiversity Issues in Open Data Sets for the Developing World, Nov

    S HANKAR , S., H ALPERN , Y., B RECK , E., A TWOOD , J., W ILSON , J., AND SCULLEY , D. No Classification without Representation: Assessing Geodiversity Issues in Open Data Sets for the Developing World, Nov

  37. [45]

    Face value: The irresistible influence of first impressions

    T ODOROV, A. Face value: The irresistible influence of first impressions. In Face V alue. Princeton University Press, 2017

    2017

  38. [46]

    W ALLBOTT , H. G. Big girls don’t frown, big boys don’t cry—Gender differences of professional actors in communicating emotion via facial expression. Journal of Nonverbal Behavior 12 , 2 (June 1988), 98–106

    1988

  39. [47]

    C., CRAWCOUR , S., W ERNER , P., A L-HAMADI , A., AND ANDRADE , A

    W ALTER , S., G RUSS , S., E HLEITER , H., T AN, J., T RAUE , H. C., CRAWCOUR , S., W ERNER , P., A L-HAMADI , A., AND ANDRADE , A. O. The biovid heat pain database data for the advancement and systematic validation of an automated pain recognition system. 2013 IEEE Internatio...

    2013

  40. [48]

    S., H U, X., AND FU, G

    W ANG , Q., C HEN , G., W ANG , Z., H U, C. S., H U, X., AND FU, G. Implicit Racial Attitudes Influence Perceived Emotional Intensity on Other-Race Faces. PLoS ONE 9 , 8 (Aug. 2014), e105946

    2014

  41. [49]

    P., AND RULE , N

    W ILSON , J. P., AND RULE , N. O. Facial trustworthiness predicts extreme criminal-sentencing outcomes. Psychological science 26 , 8 (2015), 1325–1331

    2015

  42. [50]

    P., AND RULE , N

    W ILSON , J. P., AND RULE , N. O. Hypothetical sentencing decisions are associated with actual capital punishment outcomes: The role of facial trustworthiness. Social Psychological and Personality Science 7 , 4 (2016), 331–338

    2016

  43. [51]

    Relative Uncertainty Learning for Facial Expression Recognition

    Z HANG , Y., WANG , C., AND DENG , W. Relative Uncertainty Learning for Facial Expression Recognition. In Advances in Neural Information Processing Systems (2021), M. Ranzato, A. Beygelzimer, Y . Dauphin, P. S. Liang, and J. W. Vaughan, Eds., vol. 34, Curran Associates, Inc., ...

    2021

  44. [52]

    Age progression/regression by conditional adversarial autoencoder

    Z HANG , Z., S ONG , Y., AND QI, H. Age progression/regression by conditional adversarial autoencoder. In IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2017), IEEE

    2017

  45. [2017]

    arXiv:1711.08536 [stat]

    arXiv

  46. [2018]

    arXiv:1707.04061 [cs]

    arXiv

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.