Image Cropping on Twitter: Fairness Metrics, their Limitations, and the Importance of Representation, Design, and Agency
Open accessThe DOI matches its Crossref record; 42 of the 42 references it cites were found in a registry or an index.
Abstract
Twitter uses machine learning to crop images, where crops are centered around the part predicted to be the most salient. In fall 2020, Twitter users raised concerns that the automated image cropping system on Twitter favored light-skinned over dark-skinned individuals, as well as concerns that the system favored cropping woman's bodies instead of their heads. In order to address these concerns, we conduct an extensive analysis using formalized group fairness metrics. We find systematic disparities in cropping and identify contributing factors, including the fact that the cropping based on the single most salient point can amplify the disparities because of an effect we term argmax bias. However, we demonstrate that formalized fairness metrics and quantitative analysis on their own are insufficient for capturing the risk of representational harm in automatic cropping. We suggest the removal of saliency-based cropping in favor of a solution that better preserves user agency. For developing a new solution that sufficiently address concerns related to representational harm, our critique motivates a combination of quantitative and qualitative methods that include human-centered design.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Figures















![Fig. 6 An image with multiple AI researchers (source: [78]) with multiple almost equally likely salient points. Top: the heat map and the crops based on the first and second most salient points. Bottom: saliency scores of all salient points in the images (top left to bottom right, each row is demarcated by vertical slices). Always cropping based on the top salient point amplifies this crop on the social system compared to the next best salient point.](https://figures.cochat.ai/paper-figures/f1508679af2b4e2286eebee49f4c3c91/image-cropping-on-twitter-fairness-metrics-their-limitations-and-the-importance-figure-6.png)

Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“the algorithmic audit has emerged as one of the most popular approaches to algorithmic accountability [10,66].”
From Who Audits the Auditors? Recommendations from a field scan of the algorithmic auditing ecosystem (2022) · cites this paper as [66]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“All individuals mentioned by multiple respondents have published at least one audit: Cathy O'Neil (mentioned by 9 respondents); Dr. Joy Buolamwini, Dr. Timnit Gebru, and Deborah Raji (6 respondents); Christo Wilson (3 respondents); and Dr. Rumman Chowdhury (2 respondents) [10,28,50,66,67].”
From Who Audits the Auditors? Recommendations from a field scan of the algorithmic auditing ecosystem (2022) · cites this paper as [66]
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Black Feminist Thought: Knowledge, Consciousness, and the Politics of Empowerment.
Brewer, R. M., & Collins, P. H. (1992) · Contemporary Sociology · doi:10.2307/2074808 · Cited by 19,445
Distilling the Knowledge in a Neural Network
Hinton, G., Vinyals, O., & Dean, J. (2015) · arXiv · doi:10.48550/arxiv.1503.02531 · Cited by 14,204 · cited 1×
“In order to reduce the computational cost to be able to use the model in production, Twitter leverages a combination of fisher pruning [72] and knowledge distillation [43].”
BackgroundCrenshaw, K. (2018) · Feminist Legal Theory · doi:10.4324/9780429500480-5 · Cited by 9,510
Algorithms of Oppression: How Search Engines Reinforce Racism
Noble, S. U. (2020) · doi:10.2307/j.ctt1pwt9w5 · Cited by 2,709
Learning to predict where humans look
Judd, T., Ehinger, K., Durand, F., & Torralba, A. (2009) · 2009 IEEE 12th International Conference on Computer Vision · doi:10.1109/iccv.2009.5459462 · Cited by 2,016 · cited 1×
“The model is trained on three publicly available external datasets: Borji and Itti [11], Jiang et al. [46], Judd et al. [47].”
MethodCertifying and Removing Disparate Impact
Feldman, M., Friedler, S. A., Moeller, J., Scheidegger, C., & Venkatasubramanian, S. (2015) · Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining · doi:10.1145/2783258.2783311 · Cited by 1,827 · cited 1×
“where R = 1 denotes that the most salient point (the focal point for the crop) is on the person, A represents group status [5, 33], ε is a slack value, and P (R = 1|A = a) is the probability that when two persons, one from a and one from b, are given, the auto-cropping model has the most salience on the person from a.”
BackgroundThe Chicago face database: a free stimulus set of faces and norming data
Ma, D. S., Correll, J., & Wittenbrink, B. (2015) · Behavior Research Methods · doi:10.3758/s13428-014-0532-5 · Cited by 1,788
Fairness and Abstraction in Sociotechnical Systems
Selbst, A. D., Boyd, D., Friedler, S. A., Venkatasubramanian, S., & Vertesi, J. (2019) · Proceedings of the Conference on Fairness, Accountability, and Transparency · doi:10.1145/3287560.3287598 · Cited by 1,631 · cited 2×
“Posing concerns related to image cropping purely in terms of demographic parity fails to question the "closure of predetermined,technical problems" [38] and fails to consider sociotechnical considerations, an example of the "framing trap" [66].”
BackgroundGender and Race: (What) Are They? (What) Do We Want Them To Be?
Haslanger, S. (2000) · Noûs · doi:10.1111/0029-4624.00201 · Cited by 1,564 · cited 1×
“Many critical race scholars also define race in terms of marked or perceived difference, rather than traceable ancestry in itself [42, 44, 65].”
BackgroundJiang, M., Huang, S., Duan, J., & Zhao, Q. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298710 · Cited by 729
Research works printed but not linked: 22
- Ruha Benjamin. 2019. Race After Technology: Abolitionist Tools for the New Jim Code. John Wiley & Sons.
- J.D. Biersdorfer. 2017. Fitting Your Photos on Instagram. The New York Times (2017).
- James E Bollman, Ramana L Rao, Dennis L Venable, and Reiner Eschbach. 1999. Automatic image cropping. US Patent 5,978,519.
- Joy Buolamwini and Timnit Gebru. 2018. Gender shades: Intersectional accuracy disparities in commercial gender classification. In Conference on fairness, accountability and transparency. 77-91.
- Judith Butler, Fabiana AA Jardim, Jacqueline Moraes Teixeira, and Sebastião Rinaldi. 2020. Endangered/endangering: Schematic racism and white paranoia. Educação e Pesquisa 46 (2020).
- Kate Crawford. 2017. The trouble with bias. In Conference on Neural Information Processing Systems, invited speaker.
- Terrance de Vries, Ishan Misra, Changhan Wang, and Laurens van der Maaten. 2019. Does object recognition work for everyone?. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops. 52-59.
- Malkia Devich-Cyril. 2020. Defund Facial Recognition. The Atlantic (2020).
- Ravit Dotan and Smitha Milli. 2019. Value-laden disciplinary shifts in machine learning. arXiv preprint arXiv:1912.01172 (2019). · arXiv
- Arturo Escobar. 2018. Designs for the Pluriverse. Duke University Press.
- Karen E Fields and Barbara Jeanne Fields. 2014. Racecraft: The soul of inequality in American life. Verso Trade.
- Barbara L Fredrickson and Tomi-Ann Roberts. 1997. Objectification theory: Toward understanding women's lived experiences and mental health risks. Psychology of women quarterly 21, 2 (1997), 173-206.
- Karamjit S Gill. 1990. Summary of human-centered systems research in Europe. University of Brighton, SEAKE Centre.
- Charles Hirschman. 2004. The origins and demise of the concept of race. Population and development review 30, 3 (2004), 385-415.
- Xun Huang, Chengyao Shen, Xavier Boix, and Qi Zhao. 2015. Salicon: Reducing the semantic gap in saliency prediction by adapting deep neural networks. In Proceedings of the IEEE International Conference on Computer Vision. 262-270. · doi:10.5555/2919332.2919806
- Catharine A MacKinnon. 1987. Feminism unmodified: Discourses on life and law. Harvard university press.
- Paul Mozur. 2019. One month, 500,000 face scans: How China is using AI to profile a minority. The New York Times 4 (2019).
- Laura Mulvey. 1989. Visual pleasure and narrative cinema. In Visual and other pleasures. Springer, 14-26.
- Arvind Narayanan. 2018. Translation tutorial: 21 fairness definitions and their politics. In Proc. Conf. Fairness Account- ability Transp., New York, USA, Vol. 1170.
- Ben Shneiderman and Pattie Maes. 1997. Direct manipulation vs. interface agents. interactions 4, 6 (1997), 42-61.
- Nancy A Van House. 2011. Feminist HCI meets Facebook: Performativity and social networking sites. Interacting with computers 23, 5 (2011), 422-429.
- Qian Yang. 2017. The role of design in creating machine-learning-enhanced user experience. In 2017 AAAI spring symposium series.
Cited by
Ethical and social risks of harm from Language Models
Weidinger, L., Mellor, J., Rauh, M., Griffin, C., Uesato, J., Huang, P.-S., Cheng, M., Glaese, M., Balle, B., Kasirzadeh, A., et al. (2021) · arXiv · doi:10.48550/arxiv.2112.04359 · Cited by 477
“The authors described this emphasis on the female body as perpetuating the ‘male gaze, a term used for the pervasive depiction of women as sexual objects for the pleasure of and from the perspective heterosexual men’ (Yee et al., 2021), emphasis added.”
BackgroundInfluentialCan Generative AI improve social science?
Bail, C. A. (2024) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2314021121 · Cited by 369
Can Generative AI Improve Social Science?
Bail, C. A. (2023) · doi:10.31235/osf.io/rwtzs · Cited by 369
Sociotechnical Harms of Algorithmic Systems: Scoping a Taxonomy for Harm Reduction
Shelby, R., Rismani, S., Henne, K., Moon, A., Rostamzadeh, N., Nicholas, P., Yilla-Akbari, N., Gallegos, J., Smart, A., Garcia, E., et al. (2023) · Proceedings of the 2023 AAAI/ACM Conference on AI, Ethics, and Society · doi:10.1145/3600211.3604673 · Cited by 279
“Representational harms include instances where certain social groups experience both over- and under-exposure [27], leading to unequal visibility [242].”
BackgroundWho Audits the Auditors? Recommendations from a field scan of the algorithmic auditing ecosystem
Costanza-Chock, S., Raji, I. D., & Buolamwini, J. (2022) · 2022 ACM Conference on Fairness Accountability and Transparency · doi:10.1145/3531146.3533213 · Cited by 277
Establishing Data Provenance for Responsible Artificial Intelligence Systems
Werder, K., Ramesh, B., & Zhang, R. (. (2022) · ACM Transactions on Management Information Systems · doi:10.1145/3503488 · Cited by 106
“The case of Twitter’s cropping algorithm shows a conflict in speed and consistency versus the risk of making incorrect predictions [95].”
BackgroundBias in Machine Learning: A Literature Review
Mavrogiorgos, K., Kiourtis, A., Mavrogiorgou, A., Menychtas, A., & Kyriazis, D. (2024) · Applied Sciences · doi:10.3390/app14198860 · Cited by 144
Algorithmic Auditing and Social Justice: Lessons from the History of Audit Studies
Vecchione, B., Levy, K., & Barocas, S. (2021) · Equity and Access in Algorithms, Mechanisms, and Optimization · doi:10.1145/3465416.3483294 · Cited by 51
“it was revisiting its image cropping algorithms and changing its interface to allow users more control over image previews [58; 93].”
Hagendorff, T., Bossert, L. N., Tse, Y. F., & Singer, P. (2023) · AI and Ethics · doi:10.1007/s43681-022-00199-9 · Cited by 84
“These biases show up in tasks like object detection [82], face recognition [83], image search [84], or image cropping [85].”
BackgroundRamesh, D., Kameswaran, V., Wang, D., & Sambasivan, N. (2022) · 2022 ACM Conference on Fairness Accountability and Transparency · doi:10.1145/3531146.3533237 · Cited by 47
“Technology providers have also taken steps towards increasing transparency due to a combination of user demands and self-imposed responsibility [19, 84, 117, 123].”
BackgroundInfluentialAI’s Regimes of Representation: A Community-centered Study of Text-to-Image Models in South Asia
Qadri, R., Shelby, R., Bennett, C. L., & Denton, R. (2023) · 2023 ACM Conference on Fairness Accountability and Transparency · doi:10.1145/3593013.3594016 · Cited by 85
“It also includes considering the ways in which model outputs can stabilize and scale existing regimes of representation through patterns of over- and under-visibility for populations around the globe [79, 120].”
Cruising Queer HCI on the DL: A Literature Review of LGBTQ+ People in HCI
Taylor, J., Simpson, E., Tran, A.-T., Brubaker, J. R., Fox, S. E., & Zhu, H. (2024) · Proceedings of the CHI Conference on Human Factors in Computing Systems · doi:10.1145/3613904.3642494 · Cited by 56
“These papers discuss LGBTQ+ marginalization in various contexts, foreshadowing common themes that would later take center stage in Queer HCI research.”
BackgroundRismani, S., Shelby, R., Smart, A., Jatho, E., Kroll, J., Moon, A., & Rostamzadeh, N. (2023) · Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3544548.3581407 · Cited by 33
Fairness and Bias in Multimodal AI: A SurveyPreprint
Adewumi, T., Alkhaled, L., Gurung, N., van Boven, G., & Pagliai, I. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2406.19097 · Cited by 68
Algorithmic Institutionalism: The Changing Rules of Social and Political Life
Mendonça, R. F., Almeida, V. A. F., & Filgueiras, F. D. B. (2023) · doi:10.1093/oso/9780192870070.001.0001 · Cited by 67
Kingsley, S., Sinha, P., Wang, C., Eslami, M., & Hong, J. (2022) · Proceedings of the ACM on Human-Computer Interaction · doi:10.1145/3555149 · Cited by 61
“…felt the algorithmic systems fed damaging stereotypes about marginalized demographics; two, because associating the content of marginalized demographics with harmful content is a harmful representational bias [74] (please see section 6.2 for more detailed information on demonetization harms).”
BackgroundInfluentialWikibench: Community-Driven Data Curation for AI Evaluation on Wikipedia
Kuo, T.-S., Halfaker, A., Cheng, Z., Kim, J., Wu, M.-H., Wu, T., Holstein, K., & Zhu, H. (2024) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3613904.3642278 · Cited by 54
“For example, Twitter users discovered that the platform’s image cropping algorithm favored light-skinned over dark-skinned individuals when both are in an image [95].”
BackgroundLi, R., Kingsley, S., Fan, C., Sinha, P., Wai, N., Lee, J., Shen, H., Eslami, M., & Hong, J. (2023) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3544548.3582074 · Cited by 49
“to exhibit racial bias in image previews [74]; the ImageNet Roulette art project which let people see the results of a computer vision classier on uploaded images, often showing oensive labels [19]; Portrait AI’s smartphone app that redraws photos using dierent famous painting styles, often redrawing people of dierent…”
BackgroundBeyond Skin Tone: A Multidimensional Measure of Apparent Skin Color
Thong, W. E., Joniak, P., & Xiang, A. (2023) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51070.2023.00452 · Cited by 49
“In the context of this paper, we are inspired by previous works [7,77] and focus on cropping images with two faces.”
BackgroundMethodImana, B., Korolova, A., & Heidemann, J. (2023) · Proceedings of the ACM on Human-Computer Interaction · doi:10.1145/3579610 · Cited by 40
“In the same year, Twitter provided source code of their image salience algorithm to researchers, challenging them to evaluate it for bias [17, 103].”
Benchmark suites instead of leaderboards for evaluating AI fairness
Wang, A., Hertzmann, A., & Russakovsky, O. (2024) · Patterns · doi:10.1016/j.patter.2024.101080 · Cited by 38
“102 Systems are broader than models and encompass additional components such as the model interface and how a model is actually used.”
Morales, E. (2023) · Social Media + Society · doi:10.1177/20563051231196882 · Cited by 37
“…et al., 2021), how violent content is magnified through recommendation systems (e.g., Wood, 2017), how automated accounts are created to amplify and redirect hateful speech (e.g., Uyheng et al., 2022), or how specific affordances reinforce racial and gendered violence (e.g., Yee et al., 2021).”
BackgroundFairness through Difference Awareness: Measuring Desired Group Discrimination in LLMs
Wang, A., Phan, M., Ho, D. E., & Koyejo, S. (2025) · Proceedings of the conference - Association for Computational Linguistics. Meeting · doi:10.18653/v1/2025.acl-long.341 · Cited by 35
Coordinated Flaw Disclosure for AI: Beyond Security Vulnerabilities
Cattell, S., Ghosh, A., & Kaffee, L.-A. (2024) · Proceedings of the AAAI/ACM Conference on AI Ethics and Society · doi:10.1609/aies.v7i1.31635 · Cited by 34
“Early efforts include Twitter’s algorithmic cropping bounty (Yee, Tantipongpipat, and Mishra 2021) and QueerInAI’s queer bias bounty (Dennler et al. 2023).”
Beyond Binary Gender Labels: Revealing Gender Bias in LLMs through Gender-Neutral Name Predictions
You, Z., Lee, H., Mishra, S., Jeoung, S., Mishra, A., Kim, J., & Diesner, J. (2024) · Workshop on Gender Bias in Natural Language Processing (GeBNLP) · doi:10.18653/v1/2024.gebnlp-1.16 · Cited by 34
Self-citationAuditing saliency cropping algorithms
Birhane, A., Prabhu, V. U., & Whaley, J. (2022) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv51458.2022.00158 · Cited by 32
“In order to compare the results with Twitter’s study [27], we generated a synthetic dataset of images sampled uniformly from the six race-gender ordered pairs [(BM,BF),”
BackgroundMethodResultInfluentialSpatial-Semantic Collaborative Cropping for User Generated Content
Su, Y., Cao, Y., Deng, J., Rao, F., & Wu, Q. (2024) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v38i5.28303 · Cited by 27
The Fall of an Algorithm: Characterizing the Dynamics Toward Abandonment
Johnson, N., Moharana, S., Harrington, C., Andalibi, N., Heidari, H., & Eslami, M. (2024) · ACM Conference on Fairness, Accountability, and Transparency (FAccT) · doi:10.1145/3630106.3658910 · Cited by 28
“These dialogues occurred in sites ranging from those that are more formal or compulsory for algorithm owners ( e.g., in court [87, 88, 109, 134, 163]), to those that were more informal ( e.g., on social media [53, 86, 165, 174, 183]).”
BackgroundAlgorithmic Regimes: Methods, Interactions, and Politics
Jarke, J., Prietl, B., Egbert, S., Boeva, Y., Heuer, H., & Arnold, M. (2024) · Amsterdam University Press eBooks · doi:10.1515/9789048556908 · Cited by 24
Beyond Behaviorist Representational Harms: A Plan for Measurement and Mitigation
Chien, J., & Danks, D. (2024) · The 2024 ACM Conference on Fairness Accountability and Transparency · doi:10.1145/3630106.3658946 · Cited by 22
“In order to understand the current landscape of representational harm definitions, we conducted a brief survey via snowball sampling using the keyword “representational harms” in the ACM Digital Library and in Google Scholar.”
MethodThe Risks of Machine Learning SystemsPreprint
Tan, S., Taeihagh, A., & Baxter, K. K. (2022) · arXiv (Cornell University) · doi:10.48550/arxiv.2204.09852 · Cited by 20
“Finally, even something as fundamental as an argmax function may result in biased image crops [198].”
BackgroundHenderson, M. D. (2025) · Leadership & Organization Development Journal · doi:10.1108/lodj-05-2025-0319 · Cited by 15
Unfairness in Machine Learning for Web Systems Applications
Minatel, D., dos Santos, N. R., da Silva, A. C. M., Cúri, M., Marcacini, R. M., & Lopes, A. D. A. (2023) · Brazilian Symposium on Multimedia and the Web · doi:10.1145/3617023.3617043 · Cited by 14
“While it did not improve its ML model to eliminate this bias, the company turned off this tool to propagate discrimination no longer [108].”
BackgroundDiscrimination through Image Selection by Job Advertisers on Facebook
Rao, V. N., & Korolova, A. (2023) · ACM Conference on Fairness, Accountability, and Transparency (FAccT) · doi:10.1145/3593013.3594115 · Cited by 12
“Through enumerating limitations of our study and their impact on the ability to judge the real-world impact of advertiser image selection, combined with discussion of implied identity use in ad delivery optimization to further amplify demographic skew, our work contributes to the growing chorus of researchers [3, 4, 2…”
BackgroundAddressing discretization-induced bias in demographic prediction
Dong, E., Schein, A., Wang, Z., & Garg, N. (2025) · PNAS Nexus · doi:10.1093/pnasnexus/pgaf027 · Cited by 12
“…even mentioned as a solution to argmax bias or in algorithmic fairness broadly – for example, it does not appear in a table of six solutions by Yee et al. [2021], or as a solution approach in a survey of fair machine learning [Caton and Haas, 2020], though sampling and thresholding approaches…”
BackgroundMethodInfluentialREFLECTIONS ON ARTIFICIAL INTELLIGENCE IN THE LIGHT OF THE FOUNDATIONS OF VOCATIONAL EDUCATION ;
Mossin, E. A., Martins, G. A., Gouveia, R. C., & Pantoni, R. P. (2025) · Educação em Revista · doi:10.1590/0102-469853835 · Cited by 12
Discriminação Algorítmica de Gênero: Estudo de Caso e Análise no Contexto Brasileiro
Taso, F. T. D. S., dos Reis, V. Q., & Martinez, F. V. (2023) · Anais do Workshop sobre as Implicações da Computação na Sociedade (WICS) · doi:10.5753/wics.2023.229980 · Cited by 12
Hilbert, M., Cingel, D. P., Zhang, J., Vigil, S. L., Shawcroft, J. E., Xue, H., Thakur, A., & Shafiq, Z. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.4674573 · Cited by 10
Yee, K., Sebag, A. S., Redfield, O., Eck, M., Sheng, E., & Belli, L. S. (2023) · Workshop on Trustworthy Natural Language Processing (TrustNLP) · doi:10.18653/v1/2023.trustnlp-1.10 · Cited by 9
Self-citationMcClure, E., & Wald, B. (2022) · Feminist Philosophy Quarterly · doi:10.5206/fpq/2022.3/4.14276 · Cited by 8
“24 See also Yee, Tantipongpipat, and Mishra (2021) for more details on the research Twitter did to test for and identify the bias in their recommender system.”
BackgroundThe Enshittification of the Creative Internet
Simpson, E., & Semaan, B. (2025) · Proceedings of the ACM on Human-Computer Interaction · doi:10.1145/3757670 · Cited by 7
“Twitter’s crop, which has come under fire for focusing the subjects of their crop on lighter skinned people over dark skinned people, and on women’s bodies over their faces [84], is one of several image formatting features that participants discussed as frustrating when sharing their work online 3 .”
Bias, Accuracy, and Trust: Gender-Diverse Perspectives on Large Language ModelsPreprint
Gaba, A., Wall, E., Babu, T. R., Brun, Y., Hall, K. W., & Bearfield, C. X. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2506.21898 · Cited by 7
“Conversations around fairness and harms in sociotechnical systems have been a focal point within the CSCW and broader HCI community [52, 61, 65, 70].”
Explainable Fairness in Regulatory Algorithmic AuditingPreprint
O’Neil, C., Sargeant, H., & Appel, J. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.4756637 · Cited by 7
Bypassing Guardrails: Lessons Learned from Red Teaming ChatGPT
Zhuo, T. Y., Huang, Y., Chen, C., Du, X., & Xing, Z. (2026) · ACM Transactions on Software Engineering and Methodology · doi:10.1145/3747288 · Cited by 6
“Exclusionary Norms [9, 95].”
A content-aware tool for converting videos to narrower aspect ratios
Soe, T. H., & Slavkovik, M. (2022) · ACM International Conference on Interactive Media Experiences · doi:10.1145/3505284.3529970 · Cited by 6
“Saliency prediction has been used to crop images [33].”
BackgroundMethodDaoui, A., Yamni, M., Altameem, T. A., Ahmad, M., Hammad, M. A., Pławiak, P., Tadeusiewicz, R., & El‐Latif, A. A. A. (2023) · Sensors · doi:10.3390/s23218957 · Cited by 6
“Image cropping is the process of removing unwanted parts of an image, such as the background, facial Image cropping is the process of removing unwanted parts of an image, such as the background, facial features, or other objects [36]. features, or other objects [36].”
BackgroundDigital Platforms and Algorithmic Erasure: What are the Implications?
Madland, C., & Ofosuhene, M. (2022) · The Open/Technology in Education Society and Scholarship Association Conference · doi:10.18357/otessac.2022.2.1.137 · Cited by 5
Evaluating Gender-Neutral Training Data for Automated Image Captioning
Amend, J. J., Wazzan, A., & Souvenir, R. (2021) · IEEE International Conference on Big Data (BigData) · doi:10.1109/bigdata52589.2021.9671774 · Cited by 5
Intersectional Software Engineering as a Field
Takaoka, A. J. W., Cutrupi, C. M., & Jaccheri, L. (2025) · Software · doi:10.3390/software4030018 · Cited by 4
Santo, W. L. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2410.12148 · Cited by 4
“Whether in reference to selfies, the popularization of face filters, or the automated cropping of images on social media to center faces [66], the face has become a central part of our digital experiences, with many of these technologies only now possible as a result of advances in the subfield of…”
BackgroundColonialismo, datatificação da vida e digitalização da economia no capitalismo de dados
Leão, L. F. B., Costa, G. B., Monteiro, L. M., & Santos, P. H. S. (2024) · OBSERVATÓRIO DE LA ECONOMÍA LATINOAMERICANA · doi:10.55905/oelv22n8-162 · Cited by 3
PORDE: Explaining Data Poisoning Attacks Through Visual Analytics with Food Delivery App Reviews
Lee, H., Chun, M., & Jung, H. (2023) · ACM International Conference on Intelligent User Interfaces (IUI) · doi:10.1145/3581754.3584128 · Cited by 3
Taso, F. T. D. S., dos Reis, V. Q., & Martinez, F. V. (2025) · Journal on Interactive Systems · doi:10.5753/jis.2025.5958 · Cited by 2
Sandvig, C., Aidinoff, M., Nelson, A., Vaccaro, K., Karahalios, K. G., Bhandari, E., Biddle, E. R., Armstrong, L., Matias, J. N., Eslami, M., et al. (2026) · The MIT Press eBooks · doi:10.7551/mitpress/15997.001.0001 · Cited by 2
A Probabilistic Greedy Attempt to Be Fair in Neural Team Recommendation
Loghmani, H. G., Saeedi, M., Rueda, G., Paul, E., & Fani, H. (2026) · Computational Intelligence · doi:10.1111/coin.70221 · Cited by 2
Information Extraction from Social Media: A Hands-on Tutorial on Tasks, Data, and Open Source Tools
Mishra, S., Rezapour, R., & Diesner, J. (2022) · ACM International Conference on Information and Knowledge Management (CIKM) · doi:10.1145/3511808.3557503 · Cited by 2
Self-citationMendonça, R. F., Filgueiras, F. D. B., & Almeida, V. A. F. (2023) · Oxford University Press eBooks · doi:10.1093/oso/9780192870070.003.0002 · Cited by 3
Redesigning Governments through Algorithmic Systems
Mendonça, R. F., Filgueiras, F. D. B., & Almeida, V. A. F. (2023) · Oxford University Press eBooks · doi:10.1093/oso/9780192870070.003.0005 · Cited by 2
FairDeFace: Evaluating the Fairness and Adversarial Robustness of Face Obfuscation MethodsPreprint
Khorzooghi, S. M. S. M., Thota, P., Singhal, M., Asudeh, A., Das, G., & Nilizadeh, S. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2503.08731 · Cited by 2
“Threshold: ϵ is typically set to 0.2 [59], but this value can be quite stringent, and prior works have employed different thresholds based on the context [60], [61], [62].”
InfluentialInterpretable and Causal Analysis for Multivariate Time Series
Dhaou, A. (2024) · theses.fr (ABES) · doi:10.70675/5d446b35z9b19z42afz85e4zf645a7788981 · Cited by 2
FLARE-AI: Flaw Reporting for AIPreprint
Longpre, S., Zhu, E., Ezell, C., Ghosh, A., McGregor, S., Paeth, K., Klyman, K., Kapoor, S., Bommasani, R., Appel, R. E., et al. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2606.31567 · Cited by 2
“Early efforts in the AI domain include bias bounty programs like Twitter’s algorithmic cropping bounty (Yee et al., 2021) and QueerInAI’s queer bias bounty (Dennler et al., 2023 (Anthropic, 2024).”
(2022) · doi:10.1145/3531146.3533103 · Cited by 1
“For example, after finding that its image cropping algorithm was more likely to crop out black or masculine-presenting individuals than white or feminine-presenting individuals, Twitter has planned to remove the use of algorithmic cropping altogether [Yee et al., 2021].”
BackgroundInformation Extraction from Social Media: A Hands-On Tutorial on Tasks, Data, and Open Source Tools
Mishra, S., Rezapour, R., & Diesner, J. (2022) · Lecture notes in computer science · doi:10.1007/978-3-030-99739-7_74 · Cited by 1
Self-citationMendonça, R. F., Filgueiras, F. D. B., & Almeida, V. A. F. (2023) · Oxford University Press eBooks · doi:10.1093/oso/9780192870070.003.0004 · Cited by 1
Mendonça, R. F., Filgueiras, F. D. B., & Almeida, V. A. F. (2023) · Oxford University Press eBooks · doi:10.1093/oso/9780192870070.003.0006 · Cited by 1
14. Power and Resistance in the Twitter Bias Discourse
Lopez, P. (2024) · Amsterdam University Press eBooks · doi:10.1515/9789048556908-014 · Cited by 1
Avila, J. I. C., Castillo, M. D. L. A. E., & Castillo, C. A. D. (2024) · LACCEI International Multiconference on Entrepreneurship, Innovation and Regional Development (LEIRD) · doi:10.18687/leird2024.1.1.238 · Cited by 1
The Possibilities of Data Resistance in a Digital Society
Beadle, K. (2024) · doi:10.5871/digital-society/9780856726897.001 · Cited by 1
Enhancing crime prediction with machine vision: Convolutional neural network approach
Akinmuyisitan, T. M. (2024) · Brunel University Research Archive (BURA) (Brunel University London) · Cited by 1
AI Ethics in Practice, Challenges and Limitations
John-Mathews, J.-M. (2021) · theses.fr (ABES) · doi:10.70675/13b9d1dfzf980z4d75za174za8622628db95 · Cited by 1
Machine Learning Models for Multimodal Detection of Anomalous Behaviors
de Gevigney, V. D. (2021) · theses.fr (ABES) · doi:10.70675/9432cbd2z2e11z4bf8z9a72z9b3e0fa767a1 · Cited by 1
Enjeux éthiques de l'apprentissage automatique et du Big Data : une perspective herméneutique
Demichelis, R. (2023) · theses.fr (ABES) · doi:10.70675/c644fb5azca2bz4644zbed6z0e4f6b41dfe0 · Cited by 1
Performance analysis of saliency detection in images
Reddy, M. S., & Nithya, B. (2022) · Materials Today Proceedings · doi:10.1016/j.matpr.2022.07.343
(2023) · Oxford University Press eBooks · doi:10.1093/oso/9780192870070.002.0004
Reading Contemporary Issues via the Lens of Algorithmic Institutionalism
Mendonça, R. F., Filgueiras, F. D. B., & Almeida, V. A. F. (2023) · Oxford University Press eBooks · doi:10.1093/oso/9780192870070.003.0003
Mendonça, R. F., Filgueiras, F. D. B., & Almeida, V. A. F. (2023) · Oxford University Press eBooks · doi:10.1093/oso/9780192870070.003.0001
Algorithms and Politics: From an Epistocratic View to a Democratic Approach
Mendonça, R. F., Filgueiras, F. D. B., & Almeida, V. A. F. (2023) · Oxford University Press eBooks · doi:10.1093/oso/9780192870070.003.0007
(2023) · Oxford University Press eBooks · doi:10.1093/oso/9780192870070.002.0003
(2023) · Oxford University Press eBooks · doi:10.1093/oso/9780192870070.002.0005
Final Remarks: A Needed Agenda
Mendonça, R. F., Filgueiras, F. D. B., & Almeida, V. A. F. (2023) · Oxford University Press eBooks · doi:10.1093/oso/9780192870070.003.0008
Doing Participation in the Midst of Algorithm Troubles
Meunier, A. (2025) · De Gruyter eBooks · doi:10.1515/9783035628814-008
The Current State of AI Bias Bounties: An Overview of Existing Programmes and ResearchPreprint
Kucenko, S., Dennler, N. S., & He, F. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2510.02036
“Twitter (now X) subsequently performed an internal bias analysis [29] and partially decommissioned the tool [30].”
Morales, E. (2024) · cIRcle (University of British Columbia) · doi:10.14288/1.0440613
User Agency Across the Machine Learning Pipeline
Chien, J. (2025) · eScholarship (California Digital Library)
Unequal Privacy: Auditing Demographic Bias Vulnerabilities in Visual Protection Systems
Khorzooghi, S. M. S. M., Thota, P., Singhal, M., Asudeh, A., Das, G., & Nilizadeh, S. (2026) · ACM Asia Conference on Computer and Communications Security (ASIA CCS) · doi:10.1145/3779208.3785292
“Several studies, including Yee et al. [94], Bird et al. [4], and Watkins and Chen [91], critique the use of one-size-fits-all rules, instead ad-vocating for domain-specific fairness thresholds.”
Centering Context in Computing Ethics Education: An Experience Report
Munyaka, I. N. S., Flores, A. M., & Bolds, M. (2026) · Research on Equity and Sustained Participation in Engineering, Computing, and Technology (RESPECT) · doi:10.1145/3796496.3811798
“e course already covered the work of Twier's trust and safety team on reducing bias in image cropping [17] , and students' journal entries reflected a deeper unde rstanding of why such a department maers and what its absence could mean for platform decision-making.”
The Paradox of ‘Algorithmic’ Accountability and its Implications for Design
Ramesh, D., Parthasarathy, S., & Banović, N. (2026) · ACM Conference on Fairness, Accountability, and Transparency (FAccT) · doi:10.1145/3805689.3812374
Contextual and Cultural Limitations of Text-to-Image Generators: The Case of East Asia
Kim, K. (2026) · Springer eBooks · doi:10.1007/978-3-032-21685-4_14
Human versus Computer VisionPreprint
Sirotkina, E. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2608.10181
Machine Learning in Its Social Context
Fiebrink, R., Gillies, M. F. P., & Ramos, G. (2026) · Cambridge University Press eBooks · doi:10.1017/9781108872225.015
The Anatomy of the Saliency Cropping Leviathan
Prabhu, V. U., & Birhane, A. (2026) · Cambridge University Press eBooks · doi:10.1017/9781108872225.018
Vivir gobernados : la nueva ingeniería de la subjetividad en la Era Digital
Altamirano, P. (2026) · Repositorio Digital de la UNC (National University of Cordoba)
The Drawing You Watch Is the Portrait You Keep: A Replayable Machine Portrait StudioPreprint
Rajpurohit, V. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2611.08994
Metadata
- Authors
- Yee, K., Tantipongpipat, U., & Mishra, S.
- Affiliations
- Twitter (United States) (US)
- Twitter Inc., NYC, NY, USA
- Venue
- Proceedings of the ACM on Human-Computer Interaction
- Published
- 2021-10-13
- Volume
- 5
- Issue
- CSCW2
- Pages
- 1-24
- ISSN
- 2573-0142 (ISSN-L)
- OpenAlex
- W3163680969
- Cited by
- 59
- References
- 26
- Citations by year
- 2026: 5 · 2025: 6 · 2024: 10 · 2023: 24 · 2022: 11 · 2021: 3
- FWCI
- 15.02
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- Locations
- Proceedings of the ACM on Human-Computer Interaction (publishedVersion, cc-by-nd)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
- Physical Sciences › Computer Science › Artificial Intelligence › Explainable Artificial Intelligence (XAI)
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Face recognition and analysis
- Keywords
- Twitter, skin tone bias, representational harm, user agency, human-centered design, algorithmic fairness
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.