Welcome to the November 2023 SIGCHI edition of ACM TechNews.

ACM TechNews - SIGCHI Edition is a sponsored special edition of the ACM TechNews news-briefing service focused on issues in Human Computer Interaction (HCI). This service serves as a resource for ACM-SIGCHI Members to keep abreast of the latest news in areas related to HCI and is distributed to all ACM SIGCHI members the first Tuesday of every month. ACM TechNews is a benefit of ACM membership and is distributed three times per week on Mondays, Wednesday, and Fridays to over 100,000 ACM members from over 100 countries around the world. ACM TechNews provides timely coverage of established and emerging areas of computer science, the latest trends in information technology, and related science, society, and technology news. For more information on ACM TechNews and joining the ACM, please click. The Interactions mobile app is available for free on iOS, Android, and Kindle platforms. Download it today and flip through the full-color magazine pages on your tablet or view it in a simplified low-bandwidth text mode on your phone. And be sure to check out the Interactions website, where you can access current and past articles and read the latest entries in our ever-expanding collection of blogs.

BCI Helps ALS Patient Control Devices
News-Medical.net
October 25, 2023

Tim Evans, a patient with amyotrophic lateral sclerosis (ALS) participating in a clinical trial at Johns Hopkins Medicine, has regained the ability to control home devices thanks to a brain-computer interface (BCI) and an algorithm trained to translate his brain signals into computer commands. Evans can navigate options on a communication board and control room lights, streaming TV applications, and other smart devices. The device uses electrodes to record large populations of neurons on the surface of the brain, rather than individual neurons. Johns Hopkins Medicine's Nathan Crone said, "They don't change from day to day as much, so the BCI algorithm we used for controlling the computer interface did not require recalibration or retraining for at least three months." During that period, the BCI algorithm maintained a 90% accuracy rate.

Full Article

Tool May Flag Signs of Pandemic-Related Anxiety, Depression in Healthcare Workers
NYU Langone Health NewsHub
October 24, 2023

Researchers at New York University's (NYU) Grossman School of Medicine and NYU Langone Health found that an artificial intelligence tool using natural language processing (NLP) detected distress among hospital workers during the pandemic. The tool analyzed teletherapy treatment transcripts from more than 800 healthcare workers and 820 non-healthcare workers during the first COVID-19 wave. The study found that healthcare workers who mentioned working in a hospital unit, lack of sleep, or mood issues were more likely to be diagnosed with anxiety or depression than those who did not discuss such topics. NYU Langone's Dr. Naomi M. Simon said, "These results suggest that [NLP] may one day become an effective screening tool for detecting and tracking anxiety and depression symptoms."

Full Article

Computer Scientists Program Robotic Seeing-Eye Dog to Guide the Visually Impaired
BingUNews
Stephen Folkerts
October 27, 2023

Binghamton University computer scientists have developed a robotic guide dog that could one day assist visually impaired individuals and pave the way for more affordable, efficient, and accessible seeing-eye dogs. The team incorporated a leash-tugging interface that uses reinforcement learning into a quadruped robot. This allows the user to pull the robotic dog in a particular direction and have it respond accordingly. Binghamton's Shiqi Zhang said after about 10 hours of training, "These robots are able to move around, navigating the indoor environment, guiding people, avoiding obstacles, and at the same time, being able to detect the tugs." The researchers plan to add a natural language interface and intelligent disobedience capabilities, with feedback from members of the visually impaired community helping to guide their work.

Full Article

Noise-Canceling Hearables Track Heart Rate
Interesting Engineering
Jijo Malayil
October 30, 2023

Google scientists applied audio plethysmography (APG) to active noise-canceling (ANC) hearables to track users' physiological data without adding extra sensors or losing battery life. The researchers explained in a blog that APG transmits "a low-intensity ultrasound probing signal through an ANC headphone's speakers," which induces echoes detected by feedback microphones. Two rounds of experimentation with 153 people showed the researchers that APG reliably measures heart rate and heart rate variability without being affected by changes in skin tone, insufficient earbud seals, and ear canal size.

Full Article

Smart Speaker Data May Be Used in Unexpected Ways
Washington University in St. Louis
Shawn Ballard
October 26, 2023

Umar Iqbal at Washington University in St. Louis warns that makers of smart speakers are using the appliances to gather information on their owners surreptitiously. Iqbal said he and his collaborators discovered that "Amazon uses smart speaker interaction data to infer user interests and then uses those interests to target personalized ads to the user," a practice the e-commerce company did not disclose upfront. The researchers measured the collection, usage, and sharing of Amazon Echo interaction data through an auditing framework, composing several personas with preferences in specific categories and one control persona. They intercepted network traffic and deduced data usage via ads targeted to each persona online and on the speakers, learning up to 41 advertisers synchronize or share cookies with Amazon before syncing them with 247 additional third parties.

Full Article

Online Grocery Baskets Less Varied Than In-Store Carts
Cornell Chronicle
Tom Fleischman
October 30, 2023

Cornell University's Sai Chand Chintala, Jura Liaukonyte, and Nathan Yang analyzed nearly 2 million shopping trips compiled by market research company Numerator to determine online shoppers tend to choose items with less variety, and fewer fruits and vegetables, than shoppers in physical stores. The researchers algorithmically mined three years' worth of grocery purchases from 4,388 participants, learning that Instacart baskets are more similar to each other from week to week within a given household compared to in-store carts, with more than twice as many overlapping products between successive trips to the same outlet. Meanwhile, shoppers stocked Instacart baskets with 13% fewer fresh vegetables and up to 7% fewer impulse items like candy, baked goods, and chips. The researchers suggested Instacart's "Buy it again" feature may factor into the similarity and reduced variety in online baskets. Liaukonyte said the results imply online shopping may encourage greater consumer loyalty and inertia.

Full Article

VR Software Controls Robot Proxy Through Natural Movements
News from Brown
October 26, 2023

Remote operators could virtually control robot avatars through natural movement via VRoxy software developed by researchers at Brown and Cornell universities. Users can don virtual reality (VR) headgear to manipulate the robot proxies in other locations, walking through environments and collaborating with others by pointing at objects, animating the robot's head, and making facial expressions. VRoxy maps the remote space where the operator will be and the physical space where they wish to be, rendering the latter into a three-dimensional model that is displayed in the VR headset. Blue circles on the ground represent so-called teleport links that compress the larger space into the smaller within the VR environment, enabling the user to navigate the robot to designated task areas without manual control. A live feed of the environment then activates, allowing operators to move around as if actually there while the robot mimics them.

Full Article

Sony's Access Controller for the PlayStation Aims to Make Gaming Easier for People with Disabilities
Associated Press
Barbara Ortutay
October 12, 2023

Sony's Access controller for the PlayStation is aiming to help people with disabilities more easily play video games. Developed with feedback from accessibility consultants, the Access controller can be customized based on the user's needs. Rather than design controllers that focus on a particular disability, said AbleGamers' Mark Barlet, whose organization worked with Sony and Microsoft on their accessible controllers, "It's about the experience that players need to bridge that gap between a game and a controller that's not designed for their unique presentation in the world." PlayStation's Alvin Daniel said the Access controller was designed to make it "broadly applicable," meaning it does not have to be held, is durable, and features configurable button caps and thumbsticks. Even aging gamers will benefit, because Access requires less agility and strength than standard PlayStation controllers.

Full Article

Choosing Exoskeleton Settings Like a Pandora Radio Station
University of Michigan News
October 18, 2023

Engineers at the University of Michigan (U-M), the Georgia Institute of Technology, and Google have designed a simple user interface for programming exoskeleton assistance settings. A predictor algorithm repeatedly offers assistance profile pairs that wearers will likely find comfortable; users then choose between the two, and the predictor presents another profile it thinks might be even better. The researchers evaluated the method on 14 ankle exoskeleton-wearing participants as they steadily walked about 2.3 miles per hour while limited to 50 choices. When presented with 10 randomly generated profiles, the participants selected the algorithm-suggested profiles most of the time. U-M's Elliott Rouse compares the feedback-based interface to a Pandora radio-station curator, as in both instances "We are creating a model of the user's preferences and using this model to optimize the user's experience."

Full Article

New Technology 'Game Changing' for Pregnant Women with Diabetes
University of East Anglia (U.K.)
October 5, 2023

A study by researchers at the U.K.'s University of East Anglia (UEA) found that pregnant women with type 1 diabetes should be given access to hybrid closed-loop technology that uses a smartphone algorithm to automate insulin delivery. The hybrid closed-loop system continuously adjusts insulin doses every 10 to 12 minutes based on blood sugar levels. In a study of 124 pregnant women between the ages of 18 and 45 on daily insulin therapy, half were given the hybrid closed-loop technology and the other half used insulin pumps or multiple daily injections. Said UEA's Helen Murphy, "Women who used the technology spent more time in the target range for pregnancy blood sugar levels — 68% versus 56%." The study also found pregnant women using the automated system gained 7.7 pounds less weight, were less likely to experience blood pressure complications, and required fewer antenatal clinic appointments.

Full Article

VR's Growing Role in Campus Counseling
Inside Higher Ed
Lauren Coffey
October 30, 2023

Virtual reality (VR) is expanding into campus counseling by enabling counseling for current students and helping instructors train future counselors. Institutions like the University of Central Florida (UCF) are using VR to guide students through traumatic events by simulating the sights, sounds, and smells of the experiences. More than 1,500 first responders and veterans, of whom about 10% are UCF students, have used the UCF Restores program, which uses VR technology developed in-house. Roberts Wesleyan University's on-campus clinic helps students, faculty, and outside community members overcome short-term anxiety through VR-facilitated exposure therapy. Pennsylvania State University's Carlos Zalaquett collaborated with VR video provider Mindscape Commons to create 180-degree recordings of students in various environments to help train student counselors.

Full Article

AI Copilot Enhances Human Precision for Safer Aviation
MIT News
Rachel Gordon
October 3, 2023

Researchers at the Massachusetts Institute of Technology's Computer Science and Artificial Intelligence Laboratory (CSAIL) have developed an artificial intelligence (AI) copilot that could improve aviation safety by working alongside human pilots. The Air-Guardian system is comprised of an optimization-based cooperative layer that involves attention from both humans and the AI, and liquid closed-form continuous-time neural networks to analyze incoming images and look for cause-and-effect relationships. Meanwhile, the VisualBackProp algorithm helps to ensure the system understands its attention maps by determining the focal points within an image. Unlike traditional autopilot systems that react when a safety breach occurs, Air-Guardian can detect potential risks early and make proactive decisions. CSAIL's Ramin Hasani said, "This system represents the innovative approach of human-centric AI-enabled aviation. Our use of liquid neural networks provides a dynamic, adaptive approach, ensuring that the AI doesn't merely replace human judgment, but complements it."

Full Article

Moderate Voice Speed Encourages Digital Assistant Use
Penn State News
Francisco Tutella
October 30, 2023

Researchers at Pennsylvania State University, the University of Alabama, and Texas State University investigated users' perceptions of digital assistants as helpful partners or things to control. The researchers measured the effects of altering a digital assistant's voice speed and engagement style on how likely users would be to utilize and trust the device. They had hundreds of study participants compose a personal budget, craft a personal health plan, or find a dessert recipe with a digital assistant set at various voice speeds, and with either monological or more conversational interactions. The researchers observed that people were more likely to use the digital assistant if its talking speed was moderate, while dialogue-like interactions alleviated the disadvantages of faster and slower voice speeds and made the tool appear more trustworthy.

Full Article

Digital Autism Screening Tool Shows Promise
U.S. National Institutes of Health
October 2, 2023

A U.S. National Institutes of Health (NIH)-funded study suggests the tablet-based SenseToKnow screening application could enable early detection of autism spectrum disorder (ASD). Duke University researchers designed the app to record and analyze children's responses to short films programmed to elicit various behavioral patterns, and to monitor early ASD indicators like differences in social attention, facial expressions, head movements, name responses, blink rates, and motor skills. Healthcare providers used SenseToKnow to screen 475 children ages 17 months to three years, of whom 49 subsequently received an ASG diagnosis while 98 were diagnosed with non-ASD-related developmental delays. Participants testing positive for ASD had a 40.6% likelihood of a subsequent ASD diagnosis, compared to only about 15% screened as positive with the standard parent questionnaire. The probability of a positive screen resulting in later diagnosis increased to 63.4% when SenseToKnow and the questionnaire were combined.

Full Article