An article in The Economist states that new research by MIT grad student Joy Buolamwini supports the suspicion that facial recognition software is better at processing white faces than those of other people. The bias probably arises “from the sets of data the firms concerned used to train their software,” the article suggests.
Dave Gershgorn writes for Quartz, highlighting congress’ concerns around the dangers of inaccurate facial recognition programs. He cites Joy Buolamwini’s Media Lab research on facial recognition, which he says “maintains that facial recognition is still significantly worse for people of color.”
Researchers in CSAIL are developing a steering program for drones that allows them to process uncertainty and avoid hitting objects while flying autonomously. Called Nanomap, the drone uses depth measurements to determine the safest path. “This technique creates an on the fly map that lets the drone handle uncertainty as opposed to being ready in every situation,” writes John Biggs for TechCrunch.
Skydio, an autonomous drone startup founded by a group of MIT alumni, has showcased a new drone that can lock-on, follow and record its subject, writes Lucas Matney of TechCrunch. One possible use for the device is to “launch the drone, lock onto yourself, and ski down a mountain while the R1 tracked you to the bottom while capturing 4K footage,” Matney explains.
Graduate student Joy Buolamwini tested three different face-recognition systems and found that the accuracy is best when the subject is a lighter skinned man, reports Timothy Revell for New Scientist. With facial recognition software being used by police to identify suspects, “this means inaccuracies could have consequences, such as systematically ingraining biases in police stop and searches,” writes Revell.
Skydio, a startup founded by MIT alumni, will soon begin shipping its new autonomous drone, the R1, which has the ability to lock-in and record a subject in 4K video while avoiding obstacles. “Drones that fly themselves — whether following people for outdoor self-photography, which is Skydio’s intended use, or for longer-range applications like delivery, monitoring and surveillance — are coming faster than you think,” writes Farhad Manjoo for the New York Times.
EdX has witnessed growing interest in its MicroMasters certificates, which are “online, examined and graded, credit-eligible graduate-level courses that involve about a quarter of the coursework of a traditional Masters degree,” writes Adam Gordon of Forbes. As edX CEO Prof. Anant Agarwal explains, “Learning once and working for the next 30 years is obsolete; we need to move to a world where re-skilling becomes part of the culture.”
Writing for Gizmodo, Sidney Fussell explains that a new Media Lab study finds facial-recognition software is most accurate when identifying men with lighter skin and least accurate for women with darker skin. The software analyzed by graduate student Joy Buolamwini “misidentified the gender of dark-skinned females 35 percent of the time,” explains Fussell.
CSAIL researchers have developed a new navigation method that allows drones to process less data, have faster reaction times, and dodge obstacles without creating a map of the environment they’re in, writes James Vincent of The Verge. “Because we’re not taking hundreds of measurements and fusing them together, it’s really fast,” said graduate student Peter Florence.
A study co-authored by MIT graduate student Joy Buolamwini finds that facial-recognition software is less accurate when identifying darker skin tones, especially those of women, writes Josh Horwitz of Quartz. According to the study, these errors could cause AI services to “treat individuals differently based on factors such as skin color or gender,” explains Horwitz.
A drone navigation system developed by CSAIL researchers doesn’t rely on intricate maps that show the location of obstacles, but adjusts for uncertainties, reports Martin Finucane of The Boston Globe. The system could be used in “in fields from search-and-rescue and defense to package delivery,” notes Finucane.
Speaking with Rob Verger for Popular Science, Assistant Prof. Christian Catalini explains how it’s not always clear what impacts the value of cryptocurrencies, citing “enthusiasm, hype, maybe even market manipulation.”
Speaking to Bloomberg’s Emily Chang and Selina Wang, Lecturer Luis Perez-Breva suggests that fear of AI stems from confusing it with automation. Perez-Breva explains that in his view, “we need to make better businesses that actually use this technology and AI to take advantage of the automation and create new jobs.”
Graduate student Joy Buolamwini is featured on NPR’s TED Radio Hour explaining the racial bias of facial recognition software and how these problems can be rectified. “The minimum thing we can do is actually check for the performance of these systems across groups that we already know have historically been disenfranchised,” says Buolanwini.
Wired reporter Sandy Ong highlights the work of Prof. Suranga Nanayakkara, who as a postdoc at MIT helped develop the Finger Reader, a device aimed at helping people with visual impairments read without the need for clunky hardware. The Finger Reader, “lets people read only what they’re pointing at, promising a relatively fuss-free experience, especially when out and about.”