With a brief glance at a single face, emerging facial recognition software can now categorize the gender of many men and women with remarkable accuracy.

But if that face belongs to a transgender person, such systems get it wrong more than one third of the time, according to new University of Colorado Boulder research.

“We found that facial analysis services performed consistently worse on transgender individuals, and were universally unable to classify non-binary genders,” said lead author Morgan Klaus Scheuerman, a PhD student in the Information Science department. “While there are many different types of people out there, these systems have an extremely limited view of what gender looks like.”

The study comes at a time when facial analysis technologies – which use hidden cameras to assess and characterize certain features about an individual – are becoming increasingly prevalent, embedded in everything from smartphone dating apps and digital kiosks at malls to airport security and law enforcement surveillance systems.

Previous research suggests they tend to be most accurate when assessing the gender of white men, but misidentify women of color as much as one-third of the time.

“We knew there were inherent biases in these systems around race and ethnicity and we suspected there would also be problems around gender,” said senior author Jed Brubaker, an assistant professor of Information Science. “We set out to test this in the real world.”

Researchers collected 2,450 images of faces from Instagram, each of which had been labeled by its owner with a hashtag indicating their gender identity. The pictures were then divided into seven groups of 350 images (#women, #man, #transwoman, #transman, #agender, #agenderqueer, #nonbinary) and analyzed by four of the largest providers of facial analysis services (IBM, Amazon, Microsoft and Clarifai).

Notably, Google was not included because it does not offer gender recognition services.

On average, the systems were most accurate with photos of cisgender women (those born female and identifying as female), getting their gender right 98.3% of the time. They categorized cisgender men accurately 97.6% of the time.

But trans men were wrongly identified as women up to 38% of the time.

And those who identified as agender, genderqueer or nonbinary – indicating that they identify as neither male or female – were mischaracterized 100 percent of the time.

“These systems don’t know any other language but male or female, so for many gender identities it is not possible for them to be correct,” says Brubaker.

The study also suggests that such services identify gender based on outdated stereotypes.

When Scheuerman, who is male and has long hair, submitted his own picture, half categorized him as female.

The researchers could not get access to the training data, or image inputs used to “teach” the system what male and female looks like, but previous research suggests they assess things like eye position, lip fullness, hair length and even clothing.

“These systems run the risk of reinforcing stereotypes of what you should look like if you want to be recognized as a man or a woman. And that impacts everyone,” said Scheuerman.

The market for facial recognition services is projected to double by 2024, as tech developers work to improve human-robot interaction and more carefully target ads to shoppers. Already, Brubaker notes, people engage with facial recognition technology every day to gain access to their smartphones or log into their computers.

If it has a tendency to misgender certain, already vulnerable, populations that could have grave consequences.

For instance, a match-making app could set someone up on a date with the wrong gender, leading to a potentially dangerous situation. Or a mismatch between the gender a facial recognition program sees and the documentation a person carries could lead to problems getting through airport security, says Scheuerman.

He is most concerned that such systems reaffirm notions that transgender people don’t fit in.

“People think of computer vision as futuristic, but there are lots of people who could be left out of this so-called future,” he said.

The authors say they’d like to see tech companies move away from gender classification entirely and stick to more specific labels like “long hair” or “make-up” when assessing images.

“When you walk down the street you might look at someone and presume that you know what their gender is, but that is a really quaint idea from the ’90s and it is not what the world is like anymore,” said Brubaker. “As our vision and our cultural understanding of what gender is has evolved. The algorithms driving our technological future have not. That’s deeply problematic.”

Unlike in television, where disability representation has gone up in recent years, the percentage of characters with disabilities in the top 1,200 films has hit a four-year low. Just 1.6% of the 4,445 speaking characters analyzed have a disability, according to a study by the USC Annenberg School for Communication and Journalism. Five of these films revolved around an underrepresented leading character with a disability and one showcased a leading character from the LGBT community.

“Including characters with disabilities does not happen by accident,” said Lauren Appelbaum, who leads RespectAbility’s Hollywood Inclusion efforts as the organization’s VP of communications and author of The Hollywood Disability Inclusion Toolkit. “What we see on screen influences how we act in real life, but that is dependent on filmmakers choosing to include individuals with disabilities in diverse and accurate portrayals. Thus, when just fewer than two percent of films include speaking characters with disabilities, the disability community is pretty much erased on screen. When filmmakers choose to include characters with disabilities, they can help to remove the stigmas that currently exist about interacting with individuals with disabilities.”

When the Annenberg study began tracking disability four years ago, it found 2.4% of speaking characters had disabilities, staying fairly consistent at 2.7% in 2016 and 2.5% in 2017. In each of those years, at least one film (two in 2015 and 2017) had proportional representation of characters with disabilities, compared to the US Census information. None of the films evaluated from 2018 featured proportional representation of characters with disabilities when compared to the US population.

“With more than a quarter of the US population identifying as having a disability, these numbers are dismal,” Appelbaum added. “In fact, the difference between the percentage of speaking characters with disabilities and reality in the US population is the largest difference in the inclusion crisis in film, at 25.6 (27.2% of US population versus 1.6% of speaking characters).”

More than half of the films (58) evaluated in 2018 did not include a single character (even non-speaking role) with a disability, a four-year high. Furthermore, 83 films had no female characters with a disability. This is an increase from 2017 but on par with 2015.

“You’re basically seeing the erasure of whole communities,” said Marc Choueiti, program director at the Annenberg Inclusion Initiative and one of the study’s authors.

“Entertainment contributes to our values and ideals,” RespectAbility’s president Jennifer Laszlo Mizrahi added. “With just 1.6% of speaking characters having disabilities in film, compared to 25% of American adults having a disability, we will continue to work with entertainment leaders to promote positive, accurate, diverse and inclusive media portrayals on TV and in film. Disability impacts every gender, race, age and sexual orientation. We want the film industry to understand that accurate, authentic and diverse portrayals of disability benefit everyone.”

Disability Affects All

Despite the fact that people of all races, ethnicities, genders, sexual orientations, etc., have disabilities, the films evaluated in this study do not show it.

Nearly three-quarters of the characters with disabilities were male (72.5%) and 27.5% were female. Most characters with disabilities were white (63.1%), while 36.9% were from underrepresented racial/ethnic groups. Just two characters with disabilities were LGB.

One statistic represents improvement, however. The percentage of underrepresented characters with disabilities increased 9.9% points compared to 2017. A corresponding decrease in white characters occurred. However, the numbers leave a lot of room for additional improvement.

“Once again, the predominant picture of characters with disabilities is one of a straight, white, male,” the study reports. “These results have remained consistent across the four years of films examined, which means that for nearly half of a decade, audiences have seen persistent under and misrepresentation of individuals with disabilities in top movies.”

Disabilities Represented

A total of nine films had a lead or co-lead character with a disability. These individuals experienced depression, dyslexia, disfigurement, blindness, heart conditions, HIV/AIDS or missing limbs. Five of these films centered on an underrepresented leading character with a disability, and one movie showcased the story of an LGBT leading character with a disability. In terms of ensemble casts, two films featured leading characters with disabilities, one male and one female, both of whom were white, and one was bisexual.

In terms of all characters with a speaking role, more than half of the characters (38) were shown to have a physical disability (55.1%), including mobility issues, amputation or severe disfigurement. Nearly one-third (30.4%) of characters have a cognitive disability, such as depression, anxiety or PTSD. And 27.5% of characters were shown with a communicative disability, such as blindness or deafness. As a character could experience a disability in more than one domain, the percentages do not total to 100%.

When looking behind the camera, there are no statistics for people with disabilities. Currently, no major production company tracks disability status for any of its employees, so the data does not yet exist.

Room for Improvement

While the statistics for characters with disabilities – as well as those who are LGBTQ – are lackluster, the overall numbers for women and people of color as leads increased in meaningful ways.

“The good news is companies are making more of an effort to be inclusive,” said Stacy L. Smith, director of the USC Annenberg Inclusion Initiative and one of the study’s authors. “We’re seeing movement. Of course, we always want it to be faster, but all of the activism and advocacy appears to be yielding results.”

She added, “We’re seeing that studios are recognizing that all that mythologizing about who can lead a film or carry a film was just that — mythologizing.”

This November, the UP Repertory Company opens its 47th production season with DADA: The Chika Minutes¸ a performance of LGBTQIA stories. 

DADA: The Chika Minutes is a documentary theater performance of various narratives from the LGBTQIA community, exploring documentary merged with the company’s own art form, tula dula (a poem written in verse, acted out in mime). With 16 segments of a spectra of LGBTQIA narratives, DADA will be tackling the real life experiences of kids to OFWs to pageant queens to sex workers to community builders and activists. 

“In this period of intense hate crimes, controversies and fake news about SOGIE, we bring the stories of the LGBTQIA you might have missed,” UP Repertory Company stated. 

With the dramaturgy of Ligaya Sinfuego, additional choreography by Michelle Alde, and choreography and direction by Jasper Villasis and Malvin Ramos.

DADA: The Chika Minutes will be staged this coming November 28 and 29 (7pm) and November 30 (3pm and 7pm) in UP Diliman.

For details on tickets and show venue, contact Hillary Guevara at 0906 626 0900.

Worldwide, gender nonconformity is on the rise. Children who don’t conform to their birth sex are often perceived less positively, which may harm their well-being. A new study of Chinese kindergarten- and elementary-school-age children looked at the development of biases against gender-nonconforming peers and tested an intervention to modify their biases. The study found that although children were indeed less positive toward gender-nonconforming peers than toward gender-conforming peers, showing children certain examples of gender-nonconforming peers reduced bias against them. These findings can inform efforts to reduce bias against gender nonconformity.

The study – titled “Child Development, Children’s Appraisals of Gender Nonconformity: Developmental Pattern and Intervention” – was conducted by researchers at the University of Hong Kong, the University of Toronto Mississauga, and the Chinese University of Hong Kong. It is published in Child Development, a journal of the Society for Research in Child Development.

“Our study breaks new ground by showing that 8- and 9-year-olds can be influenced to be more positive toward their gender-nonconforming peers,” notes Ivy Wong, assistant professor of psychology and gender development at the Chinese University of Hong Kong and the University of Hong Kong, who led the study. “The findings can help develop strategies to reduce bias against gender nonconformity; a bias which appears to place some children at risk of peer rejection.”

Researchers tested 315 Hong Kong Chinese children in two studies. In the first study, 4-, 5-, 8-, and 9-year-olds watched vignettes about hypothetical boys and girls who were gender conforming and gender nonconforming in play, appearance, preferences for playmates, and activities. Afterwards, they were asked questions about their views of the children in the vignettes, took part in a behavioral task about sharing with the children in the vignettes, and ranked the children they had watched from most to least favorite.

The study found that the children as young as 4-years old gave less positive appraisals and shared less generously with peers who did not conform to stereotypical expressions of gender. This finding was stronger when the children watching the vignettes were older or the children in the videos were boys.

In the second study, 8- and 9-year-olds took part in an intervention before viewing the vignettes. They saw stories about boys and girls who were portrayed as defying gender expectations in playing with toys, engaging in activities, and wearing clothing. However, the hypothetical children were also portrayed as gender conforming and thus, similar to most children of their gender in certain ways (e.g., a boy who likes to wear pink also enjoys playing basketball, a girl who likes to play with action figures also likes to jump rope) and having some positive attributes (e.g., getting good grades at school). Another group of 8- and 9-year-olds was shown stories of zoo animals, which served as a control condition. Afterwards, both groups of participating children answered the same questions and took part in the same tests as in the first study.

This study found that showing 8- and 9-year-olds examples of gender-nonconforming peers who showed positive and gender-conforming characteristics–the latter of which likely increases perceived similarity to the participants, given that most children are gender normative by definition–reduced bias against gender nonconformity.

“Our study suggests that highlighting positive attributes of individuals and qualities that gender-conforming and nonconforming children share more broadly–without highlighting whether they are conforming or nonconforming–could be helpful,” according to Doug VanderLaan, assistant professor of psychology at the University of Toronto Mississauga and co-corresponding author of the study. “For example, teachers could create opportunities for children to learn about how each person is special as an individual and ways that they are potentially similar.”

The authors acknowledge several limitations to their study, including that it was conducted in a lab and used hypothetical children in the vignettes, and that it focused on short- and not long-term effects.

TAIPEI, TAIWAN – As the country holds the 17th annual Taipei Pride parade, the transgender community of Taiwan converged in the first-ever trans Pride march. The gathering aimed to spread awareness on issues that are specific to the trans community, including the need for the Taiwanese government to remove “sexual reassignment surgery” from the list of requirements for trans people to change their gender markers.

The main event of Taipei Pride is themed “Together, Make Taiwan Better”; but to date – and akin to other LGBTQIA developments in Western countries – trans-specific issues continue to be relegated to the sidelines when discussing concerns of the LGBTQIA community, making trans-specific gatherings important.

The first trans march was held in San Francisco in 2004. And then, just as it is now, it continues to be a gathering pushing for the creation of a space for the diverse communities in the LGBTQIA community to unite and achieve the social justice and equality the LGBTQIA movement sought out in the beginning.

Pride, as a whole, has devolved into a corporate/money-earning stunt/parade; so that marches for trans and lesbians are now the more political versions.

As it is, the annual LGBTQIA Pride parade in Taiwan – first held in 2003 – attracts over 100,000 participants. Last year, 130,000 people joined the event, boosted by the passing of same-sex marriage legislation. It is the largest commercial Pride gathering in Asia, ahead of Tel Aviv Pride in Israel, which is the largest in the Middle East.

Taiwan LGBT Pride Community, the organizer of Taiwan LGBTQ Pride parade, holds the parade on the last Saturday of October.

Polyamorous families experience marginalization during pregnancy and birth, but with open, nonjudgmental attitudes from health care providers and changes to hospital policies, this can be reduced. This is according to new research – “The Polyamorous Childbearing and Birth Experiences Study (POLYBABES): a qualitative study of the health care experiences of polyamorous families during pregnancy and birth” – published in CMAJ (Canadian Medical Association Journal).

As it is, in the US alone, an estimated one in five single Americans have engaged in consensual polyamory, or consensual nonmonogamy, and interest in these types of relationships seems to be increasing. People who identify as gay, lesbian or bisexual are more likely to be in consensual nonmonogamous relationships.

Few studies exist on the experiences of polyamorous families in health care, and it appears there are none on experiences during pregnancy and birth.

“[G]iven the high proportion of polyamorous individuals who are of child-bearing age and the substantial potential for stigma, it is important to investigate polyamorous individuals’ experiences with reproductive care providers to better inform practice,” writes Dr. Elizabeth Darling, a study author and assistant dean, midwifery, and an associate professor, Department of Obstetrics & Gynecology, McMaster University, Hamilton, Ontario, with coauthors.

Several themes emerged in this qualitative study of 24 participants, including 11 women who gave birth within the last five years and 13 partners.

Key points:

• Participants deliberately planned families, choosing health care providers who they thought would be less discriminatory because of relationship status.
• More partners means more support, although some partners were not able to fully share this support because of discomfort in disclosing polyamorous relationships.
• People in polyamorous relationships often chose to disclose their status when it was medically relevant, and they received both positive and negative reactions from health care providers.
• Navigating the health system presented challenges, including administrative barriers, in which forms did not have enough space for additional partners, or newborn identification bracelets that could be issued for only two parents

To improve health care experiences for polyamorous families, the study participants suggested health care providers should acknowledge the partners’ presence and roles, be open and nonjudgmental, adapt administrative forms and procedures, and advocate for patients and their families.

“Our findings align with recent reports that individuals engaging in consensual nonmonogamy face stigma with respect to accessing health care,” write the authors. “Our results also suggest that polyamorous individuals face concerns similar to those of other gender and sexual minorities regarding administrative barriers and challenges with disclosure to health care providers.”

The authors state that substantial work needs to be done to remove marginalization experienced by these families in the health care system.

“[R]educing providers’ implicit biases toward sexual minority groups, and patients in consensually nonmonogamous relationships in particular, is vital to addressing health disparities,” writes Dr. Sharon Flicker, Department of Psychology, California State University, Sacramento, California, in a related commentary.

“Health care providers have an opportunity to mitigate this stress by providing inclusive environments and sensitive health care.”