Age Bias in Emotion Detection: An Analysis of Facial Emotion Recognition Performance on Young, Middle-Aged, and Older Adults

Kim, Eugenia; Bryant, De’Aira; Srikanth, Deepak; Howard, Ayanna M.

doi:10.1145/3461702.3462609

Cited by 51 publications

(20 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The results showed that age estimation generally performed poorly on older age groups (60 +), an effect which was compounded by gender and race; the age estimation worked disappointingly on older women of colour. Recently, another study showed that, when evaluating systems for facial emotion recognition (FER) using various classification performance metrics, the state-of-the-art commercial systems performed the best when recognizing emotions in younger adults (aged 19-31), and worst for the oldest age group (61-80) (Kim et al 2021).…”

Section: Age Bias In Algorithms and Digital Datasets (Technical Level)mentioning

confidence: 99%

AI ageism: a critical roadmap for studying age discrimination and exclusion in digitalized societies

Stypińska

2022

AI & Soc

View full text Add to dashboard Cite

In the last few years, we have witnessed a surge in scholarly interest and scientific evidence of how algorithms can produce discriminatory outcomes, especially with regard to gender and race. However, the analysis of fairness and bias in AI, important for the debate of AI for social good, has paid insufficient attention to the category of age and older people. Ageing populations have been largely neglected during the turn to digitality and AI. In this article, the concept of AI ageism is presented to make a theoretical contribution to how the understanding of inclusion and exclusion within the field of AI can be expanded to include the category of age. AI ageism can be defined as practices and ideologies operating within the field of AI, which exclude, discriminate, or neglect the interests, experiences, and needs of older population and can be manifested in five interconnected forms: (1) age biases in algorithms and datasets (technical level), (2) age stereotypes, prejudices and ideologies of actors in AI (individual level), (3) invisibility of old age in discourses on AI (discourse level), (4) discriminatory effects of use of AI technology on different age groups (group level), (5) exclusion as users of AI technology, services and products (user level). Additionally, the paper provides empirical illustrations of the way ageism operates in these five forms.

show abstract

Section: Age Bias In Algorithms and Digital Datasets (Technical Level)mentioning

confidence: 99%

AI ageism: a critical roadmap for studying age discrimination and exclusion in digitalized societies

Stypińska

2022

AI & Soc

View full text Add to dashboard Cite

show abstract

“…Wilson et al [71] finds that state-of-the-art object detection systems also fail for people with darker skin. Rhue [53] observes that emotion detection systems are more likely to ascribe negative emotions to Black individuals, while Kim et al [31] find that emotion detection systems fail to generalize for images of older adults. In accordance with this finding, Park et al [45] show that computer vision datasets systematically underrepresent older adults.…”

Section: Impact Of Training Datamentioning

confidence: 99%

American == White in Multimodal Language-and-Image AI

Wolfe

Çalışkan

2022

Proceedings of the 2022 AAAI/ACM Conference on AI, Ethics, and Society

View full text Add to dashboard Cite

Three state-of-the-art language-and-image AI models, CLIP, SLIP, and BLIP, are evaluated for evidence of a bias previously observed in social and experimental psychology: equating American identity with being White. Embedding association tests (EATs) using standardized images of self-identified Asian, Black, Latina/o, and White individuals from the Chicago Face Database (CFD) reveal that White individuals are more associated with collective in-group words than are Asian, Black, or Latina/o individuals, with effect sizes > .4 for White vs. Asian comparisons across all models. In assessments of three core aspects of American identity reported by social psychologists, single-category EATs reveal that images of White individuals are more associated with patriotism and with being born in America, but that, consistent with prior findings in psychology, White individuals are associated with being less likely to treat people of all races and backgrounds equally. Additional tests reveal that the number of images of Black individuals returned by an image ranking task is more strongly correlated with state-level implicit bias scores for White individuals (Pearson's 𝜌 = .63 in CLIP, 𝜌 = .69 in BLIP) than are state demographics (𝜌 = .60), suggesting a relationship between regional prototypicality and implicit bias. Three downstream machine learning tasks demonstrate biases associating American with White. In a visual question answering task using BLIP, 97% of White individuals are identified as American, compared to only 3% of Asian individuals. When asked in what state the individual depicted lives in, the model responds China 53% of the time for Asian individuals, but always with an American state for White individuals. In an image captioning task, BLIP remarks upon the race of Asian individuals as much as 36% of the time, but never remarks upon race for White individuals. Finally, provided with an initialization image from the CFD and the text "an American person, " a synthetic image generator (VQGAN) using the text-based guidance of CLIP lightens the skin tone of individuals of all races (by 35% for Black individuals, based on pixel brightness). The results indicate that biases equating American identity with being White are learned by language-and-image AI, and propagate to downstream applications of such models. CCS CONCEPTS• Computing methodologies → Computer vision; Natural language processing.

show abstract

“…Bias in AI. Social biases related to gender [7], age [40], religion [1], and sexuality [68] have been observed in AI systems. Our review of the extensive related work in this area focuses on racial bias in AI and on biases observed in CLIP.…”

Section: Related Workmentioning

confidence: 99%

Evidence for Hypodescent in Visual Semantic AI

Wolfe¹,

Banaji²,

Çalışkan³

2022

Preprint

View full text Add to dashboard Cite

We examine the state-of-the-art multimodal "visual semantic" model CLIP ("Contrastive Language Image Pretraining") for the rule of hypodescent, or one-drop rule, whereby multiracial people are more likely to be assigned a racial or ethnic label corresponding to a minority or disadvantaged racial or ethnic group than to the equivalent majority or advantaged group. A face morphing experiment grounded in psychological research demonstrating hypodescent indicates that, at the midway point of 1, 000 series of morphed images, CLIP associates 69.7% of Black-White female images with a Black text label over a White text label, and similarly prefers Latina (75.8%) and Asian (89.1%) text labels at the midway point for Latina-White female and Asian-White female morphs, reflecting hypodescent. Additionally, assessment of the underlying cosine similarities in the model reveals that association with White is correlated with association with "person, " with Pearson's 𝜌 as high as 0.82, 𝑝 < 10 −90 over a 21, 000-image morph series, indicating that a White person corresponds to the default representation of a person in CLIP. Finally, we show that the stereotype-congruent pleasantness association of an image correlates with association with the Black text label in CLIP, with Pearson's 𝜌 = 0.48, 𝑝 < 10 −90 for 21, 000 Black-White multiracial male images, and 𝜌 = 0.41, 𝑝 < 10 −90 for Black-White multiracial female images. CLIP is trained on Englishlanguage text gathered using data collected from an American website (Wikipedia), and our findings demonstrate that CLIP embeds the values of American racial hierarchy, reflecting the implicit and explicit beliefs that are present in human minds. We contextualize these findings within the history of and psychology of hypodescent. Overall, the data suggests that AI supervised using natural language will, unless checked, learn biases that reflect racial hierarchies. CCS CONCEPTS• Computing methodologies → Artificial intelligence.

show abstract

Age Bias in Emotion Detection: An Analysis of Facial Emotion Recognition Performance on Young, Middle-Aged, and Older Adults

Cited by 51 publications

References 12 publications

AI ageism: a critical roadmap for studying age discrimination and exclusion in digitalized societies

AI ageism: a critical roadmap for studying age discrimination and exclusion in digitalized societies

American == White in Multimodal Language-and-Image AI

Evidence for Hypodescent in Visual Semantic AI

Contact Info

Product

Resources

About