Can we teach computers to make like Mendelssohn? Duke Today, Dec. 14, 2017. Duke University researchers are teaching computers to write classical piano music in the mode of great composers like Mendelssohn and Beethoven. The resulting tunes are a pastiche of 19th century style.
Mapping electricity access for a sixth of the world’s people. Duke Research blog, Aug. 1, 2017. To get power to people who lack electricity, first officials need to locate them. But for much of the developing world, reliable, up-to-date data on electricity access is hard to come by. Now, Duke students are applying machine learning to satellite images to make the job easier.
Helping robots learn to see in 3-D. July 14, 2017. While it’s relatively straightforward for robots to “see” objects with cameras and other sensors, interpreting what they see, from a single glimpse, is difficult. New technology enables robots to spot a new object and recognize what it is, whether it is right side up or upside down, without examining it from multiple angles. It can also fill in the blind spots in its field of vision and “imagine” any parts that are hidden from view. Picked up by NPR affiliate WFDD radio.
New tools safeguard Census data about where you live and work. Duke Today, May 18, 2017. New methods enable people to learn as much as possible from Census data for policy-making and funding decisions, while guaranteeing that no one can trace the data back to your household or business. Census-related statistics are used to allocate billions of dollars annually for things like disaster relief, roads and schools. Researchers have developed algorithms that guarantee your information stays private without compromising research about your community.
Artificial intelligence meets big data. Duke Today, Oct. 11, 2016. Duke professor Cynthia Rudin is training computers to find patterns in crime, medical and other data that humans miss.
Fact-checking Senate campaign ads just got easier. Duke Today, Sept. 29, 2016. If you live in one of the battleground states in this year’s races for U.S. Senate, you have probably been inundated with political ads, many of which talk about a candidate’s willingness to toe the party line or vote across the aisle. Now, analyzing such claims for accuracy is about to get easier, thanks to a new website that lets visitors fact-check claims about congressional voting records against the data behind them.
Analog DNA circuit does math in a test tube. Duke Today, Aug. 23, 2016. Duke University researchers have created strands of synthetic DNA that, when mixed together in a test tube in the right concentrations, form an analog circuit that can add, subtract and multiply as the molecules form and break bonds. While most DNA circuits are digital, their device performs calculations in an analog fashion by measuring the varying concentrations of specific DNA molecules directly, without requiring special circuitry to convert them to zeroes and ones first. Picked up by NPR affiliate WFDD.
Is Durham’s revival pricing some longtime residents out? Duke Research Blog, Aug. 17, 2016. Durham real estate and businesses are booming. A student mapping project aims to identify the neighborhoods at risk of pricing longtime residents out. Picked up by the Durham Herald-Sun.
Of heartbeats, bones and brushstrokes. Duke Today, Aug. 1, 2016. It takes a well-trained eye to spot an irregular heartbeat in the peaks and valleys of an electrocardiogram. The same goes for identifying an extinct ape from a single fossilized tooth, or telling an original van Gogh from a fake. But in recent years, applied mathematician Ingrid Daubechies has been training computers to churn through ECG tracings, high-resolution scans of fossils, paintings and other complex digital data and work things out automatically.
Video privacy software lets you select what others can see. Duke Today, June 28, 2016. Camera-equipped smartphones, laptops and other devices make it possible to share ideas and images with anyone, anywhere, often in real-time. But in our cameras-everywhere culture, the risks of accidentally leaking sensitive information are growing. Computer scientists at Duke University have developed software that helps prevent inadvertent disclosure of trade secrets and other restricted information within a camera’s field of view by letting users specify what others can see. Picked up by the Daily Mail and NPR affiliate WFDD radio.