Researchers from MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL) believe that analyzing photos like these could help us learn recipes and better understand people's eating habits. In a new paper with the Qatar Computing Research Institute (QCRI), the team trained an artificial intelligence system called Pic2Recipe to look at a photo of food and be able to predict the ingredients and suggest similar recipes. "In computer vision, food is mostly neglected because we don't have the large-scale datasets needed to make predictions," says Yusuf Aytar, an MIT postdoc who co-wrote a paper about the system with MIT Professor Antonio Torralba. They then used that data to train a neural network to find patterns and make connections between the food images and the corresponding ingredients and recipes.
Monterrey itself has a strong incentive to take part in this study, since it loses an estimated 40 percent of its water supply to leaks every year, costing the city about $80 million in lost revenue. That's why that desert nation's King Fahd University of Petroleum and Minerals has sponsored and collaborated on much of the MIT team's work, including successful field tests there earlier this year that resulted in some further design improvements to the system, Youcef-Toumi says. Currently there is not an effective tool to locate leaks in those plastic pipes, and MIT PipeGuard's robot is the disruptive change we have been looking for." The MIT system was actually first developed to detect gas leaks, and later adapted for water pipes.
The LGPR has been used for lane keeping even when snow, fog, or dust obscures aboveground features. "Most autonomous vehicles rely on optical systems that'see' road surfaces and surrounding infrastructure to localize themselves. Even in fair conditions, having an independent sensor to rely on when your optics aren't working could add several orders of magnitude to the reliability of current autonomous lane keeping systems. Assessments of LGPR's accuracy over six- and 12-month periods show that the maps of primary roads remain valid; less stable are maps of some minor roads whose subsurfaces may be degraded by poor drainage.
MIT's initiative that brings together problem-solvers of all stripes to tackle the world's pressing problems -- has four new global challenges for 2017: brain health; sustainable urban communities; women and technology; and youth, skills, and the workforce of the future. And, it builds and convenes a community of leaders who have the resources, the expertise, the mentorship, and the know-how to get each solution piloted, scaled, and implemented. "In the two and a half years since we first announced Solve, it has evolved in important ways. The May event celebrated the first cycle of Solvers, who worked on those 2016 challenges, by bringing them together with the Solve community to form partnerships to help implement their solutions.
MIT and Harvard Medical School researchers have devised a way to image biopsy samples with much higher resolution -- an advance that could help doctors develop more accurate and inexpensive diagnostic tests. The new technique relies on an approach known as expansion microscopy, developed originally in Edward Boyden's lab at MIT, in which the researchers expand a tissue sample to 100 times its original volume before imaging it. Boyden's original expansion microscopy technique is based on embedding tissue samples in a dense, evenly generated polymer that swells when water is added. When the researchers showed the images of the expanded tissue samples to a group of scientists that included pathologists and nonpathologists, the group was able to identify the diseased tissue with 90 percent accuracy overall, compared to only 65 percent accuracy with unexpanded tissue samples.
The program connecting students and faculty with alumni and industry partners who work together to improve athletic performance by using engineering to enhance endurance, speed, accuracy, and agility in sports. Fay, who played squash and field hockey as an undergraduate at MIT, is working to identify the optimal weight for squash rackets by modeling the swing of a racket based on a person's height and weight. Cricket: Using a robotic arm to test the umpire's decisions In cricket, if a ball makes slight contact with a bat and is caught in the field before it touches the ground, the batsman is out. When the International Cricket Council (ICC) sought to test the accuracy of DRS, they turned to Sanjay Sarma, the Fred Fort Flowers and Daniel Fort Flowers Professor in Mechanical Engineering.
One of mPath's more unique recent projects was helping a toothpaste company understand people's experience with brushing their teeth. In studying customers in electronics stores, mPath found that engagement spiked while they were trying out interactive electronics, but dipped dramatically when an employee came over to deliver a so-called "sales pitch." In another instance, mPath countered a commonly held idea that letting children play a brief video game as "dessert" after reading works well. With its renewed focus on children's learning, the sensor has, in a way, come full circle: The MOXO sensor's core technology began as tool for studying stress levels of children with autism.
Standard computer chips for quadcoptors and other similarly sized drones process an enormous amount of streaming data from cameras and sensors, and interpret that data on the fly to autonomously direct a drone's pitch, speed, and trajectory. The team, led by Sertac Karaman, the Class of 1948 Career Development Associate Professor of Aeronautics and Astronautics at MIT, and Vivienne Sze, an associate professor in MIT's Department of Electrical Engineering and Computer Science, developed a low-power algorithm, in tandem with pared-down hardware, to create a specialized computer chip. The group quickly realized that conventional chip design techniques would likely not produce a chip that was small enough and provided the required processing power to intelligently fly a small autonomous drone. For each version of the algorithm that was implemented on the FPGA chip, the researchers observed the amount of power that the chip consumed as it processed the incoming data and estimated its resulting position in space.
Four years ago, researchers at MIT's Media Lab developed a computer vision system that can analyze street-level photos taken in urban neighborhoods in order to gauge how safe the neighborhoods would appear to human observers. They find that density of highly educated residents, proximity to central business districts and other physically attractive neighborhoods, and the initial safety score assigned by the system all correlate strongly with improvements in physical condition. The system that assigned the safety ratings was a machine-learning system, which had been trained on hundreds of thousands of examples in which human volunteers had rated the relative safety of streetscapes depicted in pairs of images. Similarly, the prevalence of buildings with street-facing windows also appeared to increase neighborhoods' safety scores.
The study, based on an experiment in Delhi, India, engaged preschool children in math games intended to help them grasp concepts of number and geometry, and in social games intended to help them cooperate and learn together. However, she adds, by the time the children in the study were learning formal math concepts in primary school, such as specific number symbols, the preschool intervention did not affect learning outcomes. The intervention using social games had effects on social skills but did not produce a comparable effect on math skills; the effects of the math games were specific to their math content. As the paper states, "Although the math games caused persistent gains in children's non-symbolic mathematical abilities, they failed to enhance children's readiness for learning the new symbolic content presented in primary school."