Deploying deep neural networks on mobile devices is a challenging task due to computation complexity and memory intensity. Existing works solve this problem by reducing model size using weight compression methods based on dimension reduction (i.e., SVD, Tucker decomposition and Quantization). However, the execution speed of these compressed models are still far below the real-time processing re...

Nowadays, deep learning and convolutional neural networks (CNNs) have become widespread tools in many biomedical engineering studies. CNN is an end-to-end tool which makes processing procedure integrated, but in some situations, this processing tool requires to be fused with machine learning methods to be more accurate. In this paper, a hybrid approach based on deep features extracted from Wave...

In this paper, we propose new Euler flux functions for use in a finite-volume Euler/Navier–Stokes code, which are very simple, carbuncle-free, yet have an excellent boundary-layer-resolving capability, by combining two different Riemann solvers into one based on a rotated Riemann solver approach. We show that very economical Euler flux functions can be devised by combining the Roe solver (a ful...

Abstract The ideal strength of a nano-component, which is the maximum stress of the structure, provides an insight into minute material. We conducted tensile simulations for a cylindrical-shaped Cu nano-wire composed of an atomic chain as a core wrapped around by shell(s) with the structure of (111) layers in a fcc crystal. Young’s moduli and the ideal strengths of the wires are less than a sin...

This letter presents an experimental study that shows that a 3(rd) physical dimension may be used to further increase information packing density in magnetic storage devices. We demonstrate the feasibility of at least quadrupling the magnetic states of magnetic-based data storage devices by recording and reading information from nanopillars with three magnetically-decoupled layers. Magneto-opti...

Research on inherently nanolaminated transition metal carbides is inspired by their unique properties combining metals and ceramics, such as higher damage tolerance, better machinability and lower brittleness compared to the binary counterparts, yet retaining the metallic conductivity. The interesting properties are related to their laminated structure, composed of transition-metal carbide laye...

Modulation and δ-doping strategies, in which atomically thin layers of charged dopants are precisely deposited within a heterostructure, have played enabling roles in the discovery of new physical behavior in electronic materials. Here, we demonstrate a purely structural "δ-doping" strategy in complex oxide heterostructures, in which atomically thin manganite layers are inserted into an isovale...