At the International Conference on Machine Learning, MIT researchers demonstrated an adversarial attack, fooling an AI system into thinking a 3-D printed turtle was a rifle, reports Matthew Hutson for Science. Prof. Aleksander Madry explains that the ability to fool AI systems shows, “we need to rethink all of our machine learning pipeline to make it more robust.”
In a recent blog post, Microsoft’s president and chief legal officer, Brad Smith, references research by MIT graduate student Joy Buolamwini while calling for government to regulate the use of facial recognition software. Buolamwini’s work “showed error rates of as much as 35% for systems classifying darker skinned women,” reports Dina Bass for Bloomberg.
Kate Baggaley writes for NBC News that movement tracking technology developed by MIT researchers can be helpful for monitoring the elderly or sick. The system could be used to monitor an elderly relative and, “receive an instant alert if he or she falls,” or a doctor could use it to monitor the progression of a patient’s disease, explains Baggaley.
CSAIL researchers have created a deep learning system that can isolate individual musical instruments in a video by clicking on the specific instrument, writes Andrew Liszewski for Gizmodo. The researchers suggest the system, “could be a vital tool when it comes to remixing and remastering older performances where the original recordings no longer exist,” explains Liszewski.
Katharine Schwab of Co.Design highlights graduate student Joy Buolamwini and Visiting Scholar J. Nathan Matias as “design heroes” for their commitment to keep technology fair. Schwab writes that Buolamwini has forced companies “to develop better, more equitable technology” while Matias helped “reduced the prevalence of fake news.”
This BBC Click segment highlights the artificial intelligence system developed by CSAIL researchers that can monitor people’s movements through walls. Prof. Dina Katabi explains that the device helps preserve the privacy of those being monitored by separating and encrypting, “any identifiable information from the measurement.”
Wired reporter Aarian Marshall highlights how MIT is launching a new undergraduate major that will combine computer science and urban planning. Prof. Eran Ben-Joseph explains that the motivation for the major is studying how, “you make a better connection between the training and computation, and what the implication of the work will be, for communities, for policies.”
Fast Company reporter Steven Melendez writes that CSAIL researchers have created a new system that allows a robot to detect human brainwave patterns so it knows when it made a mistake. Melendez explains that, “Teaching robots to understand human nonverbal cues and signals could make them safer and more efficient at working with people.”
In an article for The New York Times, graduate student Joy Buolamwini writes about how AI systems can often reinforce existing racial biases and exclusions. Buolamwini writes that, “Everyday people should support lawmakers, activists and public-interest technologists in demanding transparency, equity and accountability in the use of artificial intelligence that governs our lives.”
Engadget reporter Jon Fingas writes that MIT researchers have developed a tiny computer chip small enough to fit on a honeybee-sized drone that can help the drone navigate. The technology could eventually be applied to, “smart pills that navigate to where they're needed, or virtually any vehicle that may need to last for a very long time on one battery charge.”
A new system developed by MIT researchers analyzes radio signals that bounce off of human bodies to track their movement and posture from behind walls, write Saqib Shah for Fox News. Shah suggests that the system could allow military personal “to ‘see’ hidden enemies by wearing augmented reality headsets.”
CSAIL researchers have developed a new system that uses low-power radio waves to detect and track people behind walls, reports Matt Simon for Wired. The system, which can be used to detect signs of distress in elderly patients, also “distinguishes one person from another in the same way your fingerprint distinguishes you,” explains Prof. Dina Katabi.
CSAIL researchers have created a system that can sense a person’s movements through walls, writes John Biggs for TechCrunch. The system is primarily intended as a healthcare device and could help with “passive monitoring of a subject inside a room without cameras or other intrusions,” and could provide insight into disease progression, Biggs explains.
Fast Company reporter Melissa Locker writes that CSAIL researchers have developed a system that allows wireless devices to sense a person’s movement through walls. Locker explains that the technology was created as a way to help those who are elderly, as it could be used to “monitor diseases like Parkinson’s and multiple sclerosis and provide a better understanding of disease progression.”
NBC Mach reporter Tom Metcalfe writes that MIT researchers are developing autonomous boats that could be used to ferry goods and people and could help ease traffic congestion. “We believe that with fleets of very agile autonomous boats we can offload some street traffic onto the waterways,” explains Prof. Daniela Rus.