Labeling promotes infants' object categorization even when labels are rare. By 2 years, infants engage in “semi‐supervised learning” (SSL), integrating labeled and unlabeled exemplars to learn categories. However, everyday learning contexts pose substantial challenges for infants' SSL. Here, two studies (
Faces can be categorized along various dimensions including gender or race, an ability developing in infancy. Infant categorization studies have focused on facial attributes in isolation, but the interaction between these attributes remains poorly understood. Experiment 1 examined gender categorization of other‐race faces in 9‐ and 12‐month‐old White infants. Nine‐ and 12‐month‐olds were familiarized with Asian male or female faces, and tested with a novel exemplar from the familiarized category paired with a novel exemplar from a novel category. Both age groups showed novel category preferences for novel Asian female faces after familiarization with Asian male faces, but showed no novel category preference for novel Asian male faces after familiarization with Asian female faces. This categorization asymmetry was not due to a spontaneous preference hindering novel category reaction (Experiment 2), and both age groups displayed difficulty discriminating among male, but not female, other‐race faces (Experiment 3). These results indicate that category formation for male other‐race faces is mediated by categorical perception. Overall, the findings suggest that even by 12 months of age, infants are not fully able to form gender category representations of other‐race faces, responding categorically to male, but not female, other‐race faces.
more » « less- PAR ID:
- 10416135
- Publisher / Repository:
- Wiley-Blackwell
- Date Published:
- Journal Name:
- British Journal of Psychology
- Volume:
- 114
- Issue:
- S1
- ISSN:
- 0007-1269
- Page Range / eLocation ID:
- p. 71-93
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
more » « less
Abstract n = 74, 51% female, 62% non‐Hispanic White, 18% multiracial, 8% Asian, 6% Black,M age = 27.3 months, collected 2018–2020) implemented a familiarization‐novelty preference paradigm assessing 2‐year‐olds' SSL when (i) exemplars from the target category are interspersed with other objects (Study 1,d = .67) and (ii) multiple categories are learned simultaneously (Study 2,d = .74). The findings indicate 2‐year‐olds' SSL is robust enough to support object categorization despite substantial challenges posed by everyday learning contexts. -
more » « less
Abstract Experience with an object's photograph changes 9‐month‐olds’ preference for the referent object, confirming they can represent objects from pictures. However, picture‐based representations appear weaker than object‐based representations. The current study's first objective was to investigate age differences in object recognition memory after familiarization with objects' pictures. The second objective was to test whether age differences in object permanence sensitivity with picture‐based representations match those found with object‐based representations, whereby 7‐month‐olds search more for familiar hidden objects but 11‐month‐olds search more for novel ones. Six‐ and 11‐month‐olds were familiarized with an object's photo and tested on their representation of the real object by comparing their reaching for it versus a novel object. Objects were visible under conditions testing recognition memory and hidden under conditions testing object permanence. Like 9‐month‐olds, 6‐ and 11‐month‐olds preferred novelty with visible objects, showing early object recognition after picture familiarization, as well as developmental continuity. Unlike 9‐month‐olds, who switched to preferring familiarity with hidden objects, 6‐ and 11‐month‐olds switched to null preference. This pattern fails to match 7‐ and 11‐month‐olds’ hidden‐object preferences after familiarization with real objects, revealing discontinuity in sensitivity to object permanence after picture familiarization, and suggesting that picture‐based representations are weaker than object‐based ones.
-
more » « less
Abstract We examined the role of community face experience on 6‐ and 8‐month‐old Caucasian infants’ scanning of own‐ and other‐race face scanning. We measured infants’ proportional fixation time and scan path amplitudes as indices of face processing. Proportional fixation time to informationally rich face regions varied as a function of age and face race for infants living in a racially homogeneous community, whereas scan path amplitudes varied as a function of age and face race for infants living in a racially diverse community. In both communities 6‐month‐old infants did not show different responding to own‐ and other‐race faces, whereas 8‐month‐old infants responded differently to own‐ and other‐race faces. However, 8‐month‐old infants from the two communities showed different patterns of cross‐race face scanning. Therefore, experience in the community beyond the home appears to contribute to the development of differential scanning of own‐ versus other‐race faces between 6 and 8 months of age.
-
more » « less
This research examined how caregiver experience (female primary caregiver or distributed caregiving with mom and dad) influenced 10‐, 14‐, and 16‐month‐olds’ visual preferences and attention toward internal facial features of female–male face pairs, and how these behaviors related to novelty preferences in a face recognition task and speed and accuracy on a visual search task. In the visual preference task, infants visually preferred male faces, regardless of caregiver experience. Despite similarities in visual preferences, infants’ attention toward females and males’ internal facial features was related for infants with distributed caregiving only. Infants’ performance across face processing tasks most often correlated for those with female primary caregivers. Results further our understanding of how infants with female primary caregivers display specialized processing of female faces, and how infants with distributed caregiving show similarities in their attention to female and male facial features.
-
null (Ed.)Existing public face image datasets are strongly biased toward Caucasian faces, and other races (e.g., Latino) are significantly underrepresented. The models trained from such datasets suffer from inconsistent classification accuracy, which limits the applicability of face analytic systems to non-White race groups. To mitigate the race bias problem in these datasets, we constructed a novel face image dataset containing 108,501 images which is balanced on race. We define 7 race groups: White, Black, Indian, East Asian, Southeast Asian, Middle Eastern, and Latino. Images were collected from the YFCC-100M Flickr dataset and labeled with race, gender, and age groups. Evaluations were performed on existing face attribute datasets as well as novel image datasets to measure the generalization performance. We find that the model trained from our dataset is substantially more accurate on novel datasets and the accuracy is consistent across race and gender groups. We also compare several commercial computer vision APIs and report their balanced accuracy across gender, race, and age groups. Our code, data, and models are available at https://github.com/joojs/fairface.more » « less
