Modern deep neural networks are highly vulnerable to adversarial examples, which attracts more and more researchers' attention to craft powerful adversarial examples. Most of these generation algorithms create global perturbations that would affect the visual quality of adversarial examples. To mitigate such drawbacks, some attacks attempt to generate local perturbations. However, existing local adversarial attacks are time‐consuming and the generated adversarial examples are still distinguishable from clean images. In this paper, we propose a novel efficient local adversarial attack (ELAA) using model interpreters to generate severe local perturbations and improve the imperceptibly of the generated adversarial examples. Specifically, we take advantage of model interpretation methods to search the discriminative regions of clean images. Then, we generate local adversarial examples by adding masks to original clean images. We also propose a new optimization method to reduce the redundancy of local perturbations. Through extensive experiments, we show our ELAA can maintain a high attack ability while preserving the visual quality of clean images. Experimental results also demonstrate our local attack outperforms state‐of‐the‐art local attack methods under various system settings.

The objective of this paper is to observe and investigate the early evolution of the shock wave, induced by a nanosecond pulsed laser in still water. A numerical method is performed to calculate the propagation of the shock wave within 1µs, after optical breakdown, based on the Gilmore model and the Kirkwood-Bethe hypothesis. The input parameters of the numerical method include the laser pulse duration, the size of the plasma and the maximally extended cavitation bubble, which are measured utilizing a high time-resolved shadowgraph system. The calculation results are verified by shock wave observation experiments at the cavitation bubble expansion stage. The relative errors of the radiuses and the velocity of the shock wave front, reach the maximum value of 45% at 5 ns after breakdown and decrease to less than 20% within 20 ns. The high attenuation characteristics of the shock wave after the optical breakdown, are predicted by the numerical method. The quick time and space evolution of the shock wave are carefully analyzed. The normalized shock wave width is found to be independent of the laser energy and duration, and the energy partitions ratio is around 2.0 using the nanosecond pulsed laser.

The objective of this paper is to reveal the influence of different surface geometric conditions on the dynamic behavior characteristics of a laser-induced bubble collapse. A high-speed camera system was used to record the oscillation process of the laser-induced bubble on plane solid walls with different roughness and a wall containing reentrant cavities full of water or gas. The focus is on the quantitative analysis of the morphological characteristics of the cavitation bubble near the solid wall under different surface forms during the first two oscillation period. The results show that the dimensionless ratio γ, defined as the distance from the center of the bubble to the wall divided by the maximum radius of the bubble, has a great influence on the change of the cavitation shape in the direction of the vertical wall. Different surface geometries without gas in our cases have no significant effect on the collapse time of cavitation bubbles. While for the surface containing gas, the direction of movement of the bubble accompanying the micro-jet will greatly change during the collapse of the cavitation bubble, and the collapse time seems to be independent of the dimensionless ratio γ. These achievements shed the light for the engineering to avoid the damage of the micro-jet caused by design suitable surface geometry.

Orchard machinery autonomous navigation is helpful for improving the efficiency of fruit production and reducing labor costs. Path planning is one of the core technologies of autonomous navigation for orchard machinery. As normally planted in straight and parallel rows, fruit trees are natural landmarks that can provide suitable cues for orchard intelligent machinery. This paper presents a novel method to realize path planning based on computer vision technologies. We combine deep learning and the least-square (DL-LS) algorithm to carry out a new navigation line extraction algorithm for orchard scenarios. First, a large number of actual orchard images are collected and processed for training the YOLO V3 model. After the training, the mean average precision (MAP) of the model for trunk and tree detection can reach 92.11%. Secondly, the reference point coordinates of the fruit trees are calculated with the coordinates of the bounding box of trunks. Thirdly, the reference lines of fruit trees growing on both sides are fitted by the least-square method and the navigation line for the orchard machinery is determined by the two reference lines. Experimental results show that the trained YOLO V3 network can identify the tree trunk and the fruit tree accurately and that the new navigation line of fruit tree rows can be extracted effectively. The accuracy of orchard centerline extraction is 90.00%.

The objective of this paper is to reveal the attenuation characteristics of a shock wave after optical breakdown in water, with laser pulses of 10-ns duration. A high time-resolved shadowgraph method is applied to capture the temporal evolutions of the cavitation bubble wall and shock wave. The experiments are carried out on a single bubble generated far away from the free surface and the rigid walls with laser pulse energies of 22 mJ, 45 mJ and 60 mJ. The results show that a high, time-resolved, wave front velocity of the shock wave is identified, and the maximum velocity can reach up to around 4000 m/s. An asymmetric shock wave is observed at the very start of the bubble expansion stage, and the process of the sharp attenuation of wave front velocity down to sound velocity, is accomplished within 310-ns. The possible relationship of the cavitation bubble and the shock wave is discussed and a prediction model, using the maximum bubble radius and the corresponding time calculated by the Gilmore model, is proposed to calculate the location of the wave front.

Highlights Photography system built with conventional instruments realized nanosecond. Resolution nanoseconds interframe time achieved by using PIV dual-head laser nanosecond resolution. Images recorded by using double-pulse per-exposure technique.

This paper uses VAR model and select total assets of Federal Reserve, CPI, total manufacturing/industrial production, REER (real broad effective exchange rate adjusted by relative consumer prices), net exports and 3-month interbank rate as indicators to measure the impact of the US monetary policy on China, Japan, EU Area and Canada. There are several key findings. Firstly, US monetary policy shock has spillover effects on the national production, consuming price and central bank action. Secondly, the change of the value of US dollar does not have significant effect on exchange rate for China, EU Area and Japan due to the adoption of flexible exchange rate, while there is significant change in Canada. Thirdly, the monetary policy shock from US is not transmitted to national trade. Finally, the decline of real effective exchange rate in the short-run shows the increasing external competitiveness.