The results sound like one of those happy-ending reveals on an episode of Queer Eye: a transgender woman achieves the one aspect of femininity that has eluded her—an authentic voice that sounds female and matches her identity. But this breakthrough is not the work of the Fab 5. It’s the result of research from the Steinhardt School’s Biofeedback Intervention Technology for Speech Lab, led by Associate Professor Tara McAllister.
A Trans voice, McAllister says, “is a complicated issue. People [think it’s] pitch. Women talk like this” (she makes her voice high and singsong) “and men talk like this” (she says, with a deep, gruff, billy-goat voice). But the key to authenticity is actually in frequency. “To make a trans woman sound authentically feminine, there are characteristic frequency bands,” she says. “You have to make adjustments to resonating frequencies on the vocal tracks.” Which can be as tricky as it sounds; it’s no small task to change the way you’ve been doing something your entire life. “Speech is a habitual behavior,” says McAllister. “If you want to change the way you talk, it takes significant effort and a lot of practice.”
Visual-acoustic biofeedback makes practice easier. In McAllister’s proof-of-concept study, she and coauthor Deanna Kawitzky recorded trans women speaking into a microphone, letting them see the acoustic signal of their resonant frequencies on a monitor. Then the speakers were given target frequencies akin to those of cisgender female speakers. “Training was brief, maybe an hour, but they were able to learn how to interpret this visual display of their acoustic signal and were able to match a target,” she says. Then researchers called in listeners they’d recruited and asked if the voices sounded typically male or female. “We wanted to know if outside listeners would hear it as sounding more consistent with female talkers,” she says. The speech with higher resonant frequencies was indeed perceived to be more feminine.
McAllister allows that this research is just the start—she and Kawitzky foresee longer-term study in the lab (with eight to 10 weeks of training) and eventually creating an app that gives the same visual biofeedback, but for home use. In the meantime, “I’ve had a few trans women ask about an app I developed for kids,” McAllister says. “[It’s] for the R sound, so it’s not optimal. But if they want to play around with the wave display, they can use it in their own practice.” —Rory Evans
In industries ranging from jewelry to medicine to aerospace, manufacturers rely on three-dimensional printing to make products including platinum necklaces, prosthetic limbs, critical turbine engine blades, and more. And in doing so, they have built a sector that analysts expect will be worth a whopping $12 billion by the end of 2019.
That’s a lot of money—enough to be tantalizing to profiteering hackers, intellectual-property thieves, and counterfeiters. But success for a few opportunistic hackers could be deadly for many: a faulty fake aortic valve could cause a fatal heart attack or stroke; a flawed counterfeit jet engine part could result in a disastrous airplane crash.
Nikhil Gupta, an associate professor of aerospace and mechanical engineering at the Tandon School of Engineering, is leading the charge to prevent exactly that from happening.
“How do you create security in these hardware parts?” he asks. “How do you create trust in an industry that is fundamentally about mistrust?”
Gupta’s team has an answer that hinges on bar codes and Quick Response (QR) codes, like those on an airplane or concert ticket. The solution they developed is to convert the flat QR code into a 3-D form built into the layers of the material being used to make the object, whether that’s a precious metal, a polymer, or some other substance. As the object is completed, additional layers of material then hide the code from the naked eye. Hackers cannot copy what they cannot see. The security code is only readable by the end user with a CT scan, an ultrasound, or an MRI.
“You just have all the information encoded inside your part,” says Gupta, who has applied for a patent for this technology. “You create a self-certification process in such a way that whoever has the part should be able to go somewhere and get their trust from the part itself.”
To further ensure that the codes remain secure during printing, Gupta’s team devised a way for the codes to be broken up and dispersed throughout the layers of the item. Only trusted end users will know how to orient the object so its code can be read and understood; hackers will be misdirected by false faces of code that create decoys.
“The idea here is that you don’t trust anybody,” says Gupta—not hackers, or counterfeiters, or even coworkers who may become disgruntled and steal intellectual property to take to a rival company.
Mindful of the cat-and-mouse element inherent to cybersecurity and constantly searching for weaknesses in his own innovation and ways to guard against them, Gupta has a group of undergraduates trying to hack objects with 3-D QR codes written into them.
Next, Gupta is investigating how to convert design files into audio files that can then be read by an app that can scan the code and reveal if the design is intact or has been tampered with.
“It’s a different project,” he says, “but it’s the same theme of security and trustworthiness.” —Sara Ivry
As Dr. Seuss, Theodor Geisel populated his stories with delightful nonsense. Or did he? Although it may come as a super-zooper-flooper-do surprise, at least some of his brain babies were rooted in reality—and his own travels—according to a recent study by primatologist James Higham, associate professor of anthropology. The Lorax was written in 1970 while Geisel was visiting Kenya, which led to speculation that the titular orange-haired, mustachioed buzzkill/soothsayer/eco-warrior might be based on the patas monkey—as even his beloved Truffula Trees bear a striking resemblance to the region’s whistling thorn acacia.
After the idea came to the attention of Higham, it was the facial-recognition database of primates he created that provided the evidence. “I thought it was a fun and interesting piece of history,” he recalls. “An opportunity to do some engaging outreach about patas monkeys and conservation.”
Higham added an image of the Lorax into his database. When the software was run, the results were either blue or patas monkey. Higham points out that the “machine classification was really provided to add support to the idea that it is plausible that the patas monkey was the inspiration for the Lorax,” he says. “They both are orange and mustached. Even the description of the voice of the Lorax sounds like the alarm call of the patas monkey.” While the results are not meant to be definitive, they are meant to be informative. And thought-provoking: Does every species have a parallel Seuss-iverse twin? “I am not sure about facial recognition,” Higham jokes, “but based on personality, I’d probably be classified as the Grinch!” —Rory Evans
| Video: Who Was the Real Lorax? |
NYU Researchers Follow the Clues
For many women riding public transit in New York City, the toll far exceeds the fare. Everyone encounters delays, crowds, “showtime!” performances, and manspreading. But for women, there’s also groping, lewd comments, and lecherous gazes. It adds up financially, according to a recent survey by the NYU Rudin Center for Transportation. Because female-identifying riders feel unsafe or uncomfortable, personal safety and peace of mind can end up costing money. It’s the Pink Tax—the extra money women spend to get around—a median increase of $26 to $50 a month due to safety concerns (and another $100 for women who are caregivers for children and family members).
The survey was inspired by #MeToo, says Rudin Associate Director Sarah Kaufman. “The movement had been picking up steam, but I knew that aside from these high-profile cases, there were women dealing with harassment every day, outside of work. I wanted to quantify it,” says Kaufman. The survey found that 75 percent of women have experienced harassment or theft (compared with 47 percent of men) and 29 percent of women don’t take public transportation late at night (for men, it’s just 8 percent). Women make up about 75 percent of for-hire vehicle and taxi riders, interestingly the same percentage that experienced some form of harassment or theft. —Rory Evans