By making a neural-network computer model that can be fooled by optical illusions like humans, the researchers advanced knowledge of the human visual system and may help improve artificial vision.
Consumers' credit cards are declined surprisingly often in legitimate transactions. One cause is that fraud-detecting technologies used by a consumer's bank have incorrectly flagged the sale as suspicious. Now MIT researchers have employed a new machine-learning technique to drastically reduce these false positives, saving banks money and easing customer frustration.
MIT computer scientists have developed a system that learns to identify objects within an image, based on a spoken description of the image. Given an image and an audio caption, the model will highlight in real-time the relevant regions of the image being described.
An ORNL research team led by Jay Jay Billings has continuously updated a workflow management system they first developed in 2010 to help computational scientists develop software, visualize data, and solve problems, saving time and effort expended in support of modeling and simulation experiments. Recently, the team published an article inSoftwareX that both details the history of the system and previews the potential benefits of upcoming versions.
New plume dispersal model alerts authorities how much time people have to flee after an attack.
More and more data and images are generated during ocean research. In order to be able to evaluate the image data scientifically, automated procedures are necessary. Together with GEOMAR data management, researchers at GEOMAR have now developed a standardized workflow for sustainable marine image analysis for the first time and recently published it in the international journal Scientific Data. Dr. Timm Schoening, lead author, will present these and other methods of digitisation in marine research at the Digital Week in Kiel.
UCLA Samueli engineers have developed a new tool to model how magnetic materials, which are used in smartphones and other communications devices, interact with incoming radio signals that carry data. It accurately predicts these interactions down to the nanometer scales required to build state-of-the-art communications technologies.
Graphene is considered a promising candidate for the nanoelectronics of the future. In theory, it should allow clock rates up to a thousand times faster than today's silicon-based electronics. Scientists from the Helmholtz Zentrum Dresden-Rossendorf (HZDR), the University of Duisburg-Essen (UDE), and the Max Planck Institute for Polymer Research (MPI-P), have now shown that graphene can actually convert electronic signals with frequencies in the gigahertz range extremely efficiently into signals with several times higher frequency.
Physicists employ a version of Maxwell's demon to reduce entropy in a three-dimensional lattice of super-cooled, laser-trapped atoms, a process that could help speed progress toward creating quantum computers.
The 'chicken or egg' paradox was first proposed by philosophers in Ancient Greece to describe the problem of determining cause-and-effect. Now, a team of physicists from The University of Queensland and the NÉEL Institute has shown that, as far as quantum physics is concerned, the chicken and the egg can both come first.