ABSTRACT
Recent evidence shows that AI-generated faces are now indistinguishable from human faces. However, algorithms are trained disproportionately on White faces, and thus White AI faces may appear especially realistic. In Experiment 1 (N = 124 adults), alongside our reanalysis of previously published data, we showed that White AI faces are judged as human more often than actual human faces-a phenomenon we term AI hyperrealism. Paradoxically, people who made the most errors in this task were the most confident (a Dunning-Kruger effect). In Experiment 2 (N = 610 adults), we used face-space theory and participant qualitative reports to identify key facial attributes that distinguish AI from human faces but were misinterpreted by participants, leading to AI hyperrealism. However, the attributes permitted high accuracy using machine learning. These findings illustrate how psychological theory can inform understanding of AI outputs and provide direction for debiasing AI algorithms, thereby promoting the ethical use of AI.
Subject(s)
Algorithms , Machine Learning , Adult , HumansABSTRACT
For decades, psychology has relied on highly standardized images to understand how people respond to faces. Many of these stimuli are rigorously generated and supported by excellent normative data; as such, they have played an important role in the development of face science. However, there is now clear evidence that testing with ambient images (i.e., naturalistic images "in the wild") and including expressions that are spontaneous can lead to new and important insights. To precisely quantify the extent to which our current knowledge base has relied on standardized and posed stimuli, we systematically surveyed the face stimuli used in 12 key journals in this field across 2000-2020 (N = 3374 articles). Although a small number of posed expression databases continue to dominate the literature, the use of spontaneous expressions seems to be increasing. However, there has been no increase in the use of ambient or dynamic stimuli over time. The vast majority of articles have used highly standardized and nonmoving pictures of faces. An emerging trend is that virtual faces are being used as stand-ins for human faces in research. Overall, the results of the present survey highlight that there has been a significant imbalance in favor of standardized face stimuli. We argue that psychology would benefit from a more balanced approach because ambient and spontaneous stimuli have much to offer. We advocate a cognitive ethological approach that involves studying face processing in natural settings as well as the lab, incorporating more stimuli from "the wild".
Subject(s)
Facial Recognition , Databases, Factual , Emotions , Facial Expression , Humans , Problem Solving , Surveys and QuestionnairesABSTRACT
BACKGROUND AND HYPOTHESIS: Schizophrenia spectrum disorders (SSD) are associated with face perception impairments. It is unclear whether impairments are equal across aspects of face perception or larger-indicating a differential impairment-for perceiving emotions relative to other characteristics (eg, identity, age). While many studies have attempted to compare emotion and non-emotion face perception in SSD, they have varied in design and produced conflicting findings. Additionally, prior meta-analyses on this topic were not designed to disentangle differential emotion impairments from broader impairments in face perception or cognition. We hypothesize that SSD-related impairments are larger for emotion than non-emotion face perception, but study characteristics moderate this differential impairment. STUDY DESIGN: We meta-analyzed 313 effect sizes from 104 articles to investigate if SSD-related impairments are significantly greater for emotion than non-emotion face perception. We tested whether key study characteristics moderated these impairments, including SSD severity, sample intelligence matching, task difficulty, and task memory dependency. STUDY RESULTS: We found significantly greater impairments for emotion (Cohen's dâ =â 0.74) than non-emotion face perception (dâ =â 0.55) in SSD relative to control samples, regardless of SSD severity, intelligence matching, or task difficulty. Importantly, this effect was obscured when non-emotion tasks used a memory-dependent design. CONCLUSIONS: This is the first meta-analysis to demonstrate a differential emotion impairment in SSD that cannot be explained by broader impairments in face perception or cognition. The findings also underscore the critical role of task matching in studies of face perception impairments; to prevent confounding influences from memory-dependent task designs.
ABSTRACT
The Duchenne marker-crow's feet wrinkles at the corner of the eyes-has a reputation for signaling genuine positive emotion in smiles. Here, we test whether this facial action might be better conceptualized as a marker of emotional intensity, rather than genuineness per se, and examine its perceptual outcomes beyond smiling, in sad expressions. For smiles, we found ratings of emotional intensity (how happy a face is) were unable to fully account for the effect of Duchenne status (present vs. absent) on ratings of emotion genuineness. The Duchenne marker made a unique direct contribution to the perceived genuineness of smiles, supporting its reputation for signaling genuine emotion in smiling. In contrast, across 4 experiments, we found Duchenne sad expressions were not rated as any more genuine or sincere than non-Duchenne ones. The Duchenne marker did however make sad expressions look sadder and more negative, just like it made smiles look happier and more positive. Together, these findings argue the Duchenne marker has an important role in sad as well as smiling expressions, but is interpreted differently in sad expressions (contributions to intensity only) compared with smiles (emotion genuineness independently of intensity). (PsycInfo Database Record (c) 2022 APA, all rights reserved).
Subject(s)
Emotions , Facial Expression , Happiness , Humans , Sadness , Smiling/psychologyABSTRACT
BACKGROUND: Despite a large body of data suggesting that delivery of fertility care to cancer patients is inconsistent and frequently insufficient, there is a paucity of literature examining training in fertility preservation for those physicians expected to discuss options or execute therapy. The study objective was to compare fertility preservation training between Reproductive Endocrinology & Infertility (REI) and Gynecologic Oncology (GYN ONC) fellows and assess the need for additional education in this field. METHODS: A 38-item survey was administered to REI and GYN ONC fellows in the United states in April 2014. Survey items included: 1) Clinical exposure, perceived quality of training, and self-reported knowledge in fertility preservation; 2) an educational needs assessment of desire for additional training in fertility preservation. RESULTS: Seventy-nine responses were received from 137 REI and 160 GYN ONC fellows (response rate 27%). REI fellows reported seeing significantly more fertility preservation patients and rated their training more favorably than GYN ONC fellows (48% of REI fellows versus 7% of GYN ONC fellows rated training as 'excellent', p < 0.001). A majority of all fellows felt discussing fertility preservation was 'very important' but fellows differed in self-reported ability to counsel patients, with 43% of REI fellows and only 4% of GYN ONC fellows able to counsel patients 'all the time' (p = 0.002). Seventy-six percent of all fellows felt more education in fertility preservation was required, and 91% felt it should be a required component of fellowship training. CONCLUSION: Significant variability exists in fertility preservation training for REI and GYN ONC fellows, with the greatest gap seen for GYN ONC fellows, both in perceived quality of fertility preservation training and number of fertility preservation patients seen. A majority of fellows in both disciplines support the idea of a standardized multi-disciplinary curriculum in fertility preservation.
ABSTRACT
A prior study of pancreatic duct-arteriovenous relationships suggested that finding ducts near muscularized blood vessels without intervening pancreatic acini indicated adenocarcinoma was present. Because focal changes of chronic pancreatitis are often seen at autopsy, it seemed reasonable to use the autopsy to test the hypothesis that this finding might be nonspecific. An unselected, consecutive series of 81 adult decedent pancreases without known pancreas cancer was evaluated for the presence of ducts near muscularized blood vessels, for fibrosis and/or atrophy, for chronic inflammation, and for duct reduplication and/or proliferation. Autolysis precluded assessment of 26% of the cases. Of evaluable cases, 37% displayed ducts near muscularized blood vessels without intervening pancreatic acini, 23% chronic inflammation, 62% fibrosis and/or atrophy, and 55% duct reduplication and/or proliferation. The finding of ducts near muscularized blood vessels was closely associated with fibrosis and/or atrophy (odds ratio = 28.87, chi = 14.59, P = 0.0001), with duct reduplication and/or proliferation (odds ratio = 19.23, chi = 15.88, P = 0.0001), but not with chronic inflammation (odds ratio = 1.41, chi = 0.05, P > 0.30). Because changes of chronic pancreatitis are associated with ducts near muscularized blood vessels and because chronic pancreatitis can mimic pancreas cancer, care should be exercised when using the finding of ducts near muscularized blood vessels without intervening pancreatic acini as a criterion for the diagnosis of pancreas cancer.
Subject(s)
Pancreas/pathology , Pancreatic Ducts/pathology , Pancreatitis/pathology , Adult , Arteries/pathology , Chronic Disease , Humans , Pancreatic Ducts/blood supply , Pancreatic Neoplasms/pathology , Veins/pathologyABSTRACT
Pulmonary thromboembolism is an often fatal complication of venous thrombosis. Any component or combination of the components composing Virchow's triad (venostasis, hypercoagulability, and endothelial damage) increases the propensity for a thrombophilic state. Hypercoagulable states may be inherited or acquired. While the etiology in many cases may be evident either on physical examination or on evaluation of the decedent's medical history, this is often not the case. We conducted a retrospective study of cases presenting to the Jefferson County Coroner/Medical Examiner's Office in Birmingham, Alabama, who were given a diagnosis of pulmonary embolism. A search of cases within the past 23 years yielded 81 cases. An underlying cause was determined in 70 cases (86%). The remaining 11 (14%) cases had no identifiable cause. We believe that a number of these cases may represent an underlying thrombophilic disorder. Since these disorders may be of an inherited or acquired nature, the determination of an etiology may be relevant to the decedent's family. Postmortem blood analyses may in selected cases be useful and appropriate for the detection of some of these disorders. However, such analyses are not practical in all cases, with each case having to be evaluated on its own merits.
Subject(s)
Pulmonary Embolism/etiology , Pulmonary Embolism/mortality , Adolescent , Adult , Age Distribution , Aged , Aged, 80 and over , Alabama/epidemiology , Black People/statistics & numerical data , Confounding Factors, Epidemiologic , Female , Forensic Medicine , Humans , Male , Middle Aged , Retrospective Studies , Risk Factors , White People/statistics & numerical dataABSTRACT
Sudden unexplained death in epilepsy occurs when epilepsy patients die suddenly and unexpectedly in the absence of recent tonic-clonic seizure activity. There is currently no known reliable indicator of acutely lethal seizure activity. Clinical studies record a relationship between recent (within 10-40 minutes) seizure activity and elevated serum prolactin levels, and postictal elevation of prolactin within peripheral vessels has proved clinically useful in determining recent seizure activity. The authors hypothesized that elevated prolactin could be detected in cerebral vessels by immunohistochemical stains, serving as a marker for sudden unexplained death in epilepsy. They conducted a retrospective study of individuals who died in their jurisdiction during the 14 years from 1986 through 1999. The study contained one group of individuals who died of sudden unexplained death in epilepsy, a group with epilepsy who died of some other cause, and a control group whose members died rapidly of a gunshot wound of the torso. Sections of hippocampus and neocortex were obtained and stained with a polyclonal prolactin antibody. No significant difference in the level of immunostaining for prolactin in cerebral vessels was found between the experimental and control groups. A review of the protocols used indicates that revision of certain aspects may provide better immunostaining and more conclusive results.