Phys.org September 5, 2019 An international team of researchers (UK, China) used a nanowire whose material composition is varied along its length, enabling it to be responsive to different colours of light across the visible spectrum. They created a series of light-responsive sections on this nanowire. Individual responses from the nanowire sections can then be directly fed into a computer algorithm to reconstruct the incident light spectrum. Every pixel of the device contains data points from across the visible spectrum, providing detailed information. This can tell us, for instance, about chemical processes occurring in the frame of the image. One […]
Novel math could bring machine learning to the next level
EurekAlert September 2, 2019 Before the neural network can begin to perform facial recognition, it is typically necessary to present it with thousands of faces. Much of these machines have been increasingly successful at pattern recognition but what goes on inside them as they learn their task is unknown. An international team of researchers (Portugal, Italy) has shown that artificial vision machines can learn to recognize complex images faster by using topological data analysis which was developed 25 years ago. Current neural networks are not good at topology. The team mathematically describe how to enforce certain symmetries, and this provides […]
Performance of electric solid propellant
Science Daily September 3, 2019 Electric solid propellants are being explored for use in dual-mode rocket engines because they aren’t susceptible to ignite from a spark or flame and can be turned on and off electrically. A team of researchers in the US (Missouri University of Science and Technology, University of Illinois, NASA Marshall Space Flight Center) compares a high-performance electric propellant with polytetrafluoroethylene (PTFE), a traditional propellant used in ablative pulsed plasma thrusters. They found that the electric solid propellant ablates about two times more than the traditional propellant. According to their analyses the physics of propellant ablation is […]
Single atoms as catalysts
Nanowerk September 2, 2019 Researchers in Austria have shown that metal atoms can be placed on a metal oxide surface so that they show exactly the desired chemical behavior. When metal atoms are deposited on a metal oxide surface, they usually have a very strong tendency to clump together and form nanoparticles. Instead of attaching the active metal atoms to a surface, it is also possible to incorporate them into a molecule with cleverly selected neighboring atoms. The molecules and reactants are then dissolved into a liquid where the chemical reactions take place. They have developed a technique to incorporate […]
This Is How Nuclear Winter Would Affect Every Single One of Us Across The Planet
Science Alert September 1, 2019 According to a team of researchers in the US (Rutgers University, NCAR, University of Colorado) if all of Russia and the US’s nuclear weapons were used in a conflict today, we could expect a shocking drop in global temperatures, less precipitation, and a lot less food to go around, nuclear particles would be transported between the hemispheres within two weeks. Global temperatures would then plunge by around 9 degrees Celsius over the next 12 months, plant growth would be limited, aerosols in the atmosphere could cause an average 30 percent drop in precipitation around the […]
Using nature to produce a revolutionary optical material
Nanowerk September 5, 2019 An international team of researchers (China, Israel, Ireland) reports on the unique nano-photonic properties of elemental tellurium particles [Te(0)], as harvest from a culture of a tellurium-oxyanion respiring bacteria. These nano-crystals prove effective in the photonic applications in the mid-infrared range compared to the chemically-formed nano-materials, suggesting a unique and environmentally friendly route of synthesis. They used the nanocrystals and a polymer to build an electro-optic switch that is immune to damage from lasers. The new material could be used to safeguard drones, surveillance cameras and other equipment against laser attacks, which can disable or destroy […]
Top 10 Science and Technology Inventions for the Week of August 30, 2019
01. Entangling photons generated millions of miles apart 02. All-optical diffractive neural network closes performance gap with electronic neural networks 03. Researchers reveal ultra-fast bomb detection method that could upgrade airport security 04. Switching electron properties on and off individually 05. Ultrathin 3-D-printed films convert energy of one form into another 06. Optical neural network could lead to intelligent cameras 07. AI at the Speed of Light 08. Graphene-based fabric protects against mosquitoes 09. Compact water harvester sucks H20 from the air 10. DARPA uses nature as a muse for new computing model And others… All Engineering Knowledge Has an […]
AI at the Speed of Light
IEEE Spectrum August 29, 2019 Researchers in Hong Kong have demonstrated a fully functioning all-optical neural network (AONN) in which linear operations are programmed by spatial light modulators and Fourier lenses, while nonlinear optical activation functions are realized in laser-cooled atoms with electromagnetically induced transparency. Because all errors from different optical neurons are independent, it is possible to scale up the size of the AONN. The hardware system is reconfigurable for different applications without the need to modify the physical structure. The AONN scheme can be used to construct various ANN architectures with intrinsic optical parallel computation…read more. Open Access […]
All Engineering Knowledge Has an Expiration Date. The Trick Is to Know When
IEEE Spectrum August 23, 2019 According to the author as new knowledge accumulates, some old knowledge becomes irrelevant and falls off the knowledge stack. Almost all the college course work one took long ago is now useless in itself, although what remains is an engineering mind-set and a mathematical grounding. The purging of obsolete knowledge is probably insufficient to make room for the new stuff, as there seems to be an exponential increase in knowledge. The complexity of our work is always increasing, similar to the increase in entropy decreed by the second law of thermodynamics. Even as Moore’s Law […]
All-optical diffractive neural network closes performance gap with electronic neural networks
Science Daily August 13, 2019 Optical computing provides unique opportunities in terms of parallelization, scalability, power efficiency, and computational speed and has attracted major interest for machine learning. Researchers at UCLA have demonstrated systematic improvements in diffractive optical neural networks, based on a differential measurement technique that mitigates the strict nonnegativity constraint of light intensity. Using this differential detection scheme, involving 10 photodetector pairs behind 5 diffractive layers with a total of 0.2 million neurons, they numerically achieved blind testing accuracies of 98.54%, 90.54%, and 48.51% for MNIST, Fashion-MNIST, and grayscale CIFAR-10 datasets, respectively. By utilizing the inherent parallelization capability […]