This week’s list of top data news highlights covers October 8, 2022 to October 14, 2022 and includes articles on meeting with coworkers in virtual reality and generating recipes with an AI system.
Microsoft has announced plans to add an app featuring the DALL-E 2 text-to-image system to its suite of Office products. Users can submit prompts to the system and add the resulting images to PowerPoint presentations, social media posts, and more. The company additionally plans to integrate the system into its Bing search engine.
Researchers at Stanford University in California have created a robotic boot that can help users walk faster while expending less energy. The boot contains sensors that collect data on force and ankle motion and a machine learning model that directs the boot to help push users’ feet off the ground during each step.
Yummy, a meal-kit company based in Estonia, has created an AI system that can turn user-submitted texts into recipe suggestions and images of the meal. Users can enter prompts related to tastes, budget, and health on their platform. The platform’s AI system then generates recipes, grocery and meal plans, and a photograph of the meals.
Legionarius, a U.S.-based medical device company, has created a smart shirt that can detect if military personnel have received an injury. The shirt contains sensors that collect data on significant tears in the fabric and relay the data to a device on the user’s back. The device can then transmit information on the location of the wound, the time of injury, and current vital signs to the user’s commanders.
Researchers with the Global Emancipation Network, a U.S.-based counter-human trafficking organization, and North Carolina State University have created two AI systems that can determine the likelihood that a massage business engages in human trafficking. The team first interviewed government officials, researchers, and law enforcement agencies to identify variables common to businesses engaged in human trafficking. They then trained the systems with public data related to the variables, such as proximity to certain businesses.
6. Updating Data-Sharing Networks
The Department of Energy (DOE) has updated its data-sharing network to better process, share, analyze, and store data across laboratories and research facilities. The new network can transfer data between 400 gigabits and 1 terabit per second and contains improved platforms for users to customize network services.
7. Detecting Mental Health Issues
Researchers at the University of Washington have created an AI system that can detect signs of decline in mental health in text messages. The team trained the system to recognize signs of cognitive distortions, such as catastrophizing in text messages. In tests, the system performed as accurately as human psychiatrists.
8. Collaborating in Virtual Reality
Microsoft and Meta have partnered to create virtual reality versions of Microsoft’s Teams, Office, Windows, and Xbox services. Workers and gamers with access to Microsoft products can use Meta’s Quest VR headsets to meet with coworkers or other gamers in virtual reality.
Researchers at Cornell University in New York have created an AI system that can predict volleyball players’ game movements, such as spiking or running, with 80 percent accuracy and determine a player’s position with nearly 85 percent accuracy. The team trained the system with videos of volleyball games.
Researchers at the University of California, San Diego have built a four-legged robot that can navigate complex terrain. The robot contains sensors and a camera that collect data on its surrounding environment and an AI system that can direct the robot around any obstacles. In many older models, robots’ AI systems use data from a single source to direct its actions. This robot’s AI system improves upon those models by combining data from its cameras and sensors.
Image credit: Flickr user JLS Photography – Alaska