Automated extraction of baleen whale calls based on the pseudo-Wigner-Ville distribution.

Baleen whales produce a wide variety of frequency-modulated calls. Extraction of the time-frequency (TF) structures of these calls forms the basis for many applications, including abundance estimation and species recognition. Typical methods to extract the contours of whale calls from a spectrogram are based on the short-time Fourier transform and are, thus, restricted by a fixed TF resolution. Considering the low-frequency nature of baleen whale calls, this work represents the contours using a pseudo-Wigner-Ville distribution for a higher TF resolution at the cost of introducing cross terms. An adaptive threshold is proposed followed by a modified Gaussian mixture probability hypothesis density filter to extract the contours. Finally, the artificial contours, which are caused by the cross terms, can be removed in post-processing. Simulations were conducted to explore how the signal-to-noise ratio influences the performance of the proposed method. Then, in experiments based on real data, the contours of the calls of three kinds of baleen whales were extracted in a highly accurate manner (with mean deviations of 5.4 and 0.051 Hz from the ground-truth contours at sampling rates of 4000 and 100 Hz, respectively) with a recall of 75% and a precision of 78.5%.

Implementing selective edge enhancement in nonlinear optics.

Recently, it has been demonstrated that a nonlinear spatial filter using second harmonic generation can implement a visible edge enhancement under invisible illumination, and it provides a promising application in biological imaging with light-sensitive specimens. But with this nonlinear spatial filter, all phase or intensity edges of a sample are highlighted isotropically, independent of their local directions. Here we propose a vectorial one to cover this shortage. Our vectorial nonlinear spatial filter uses two cascaded nonlinear crystals with orthogonal optical axes to produce superposed nonlinear vortex filtering. We show that with the control of the polarization of the invisible illumination, one can highlight the features of the samples in special directions visually. Moreover, we find the intensity of the sample arm can be weaker by two orders of magnitude than the filter arm. This striking feature may offer a practical application in biological imaging or microscopy, since the light field reflected from the sample is always weak. Our work offers an interesting way to see and emphasize the different directions of edges or contours of phase and intensity objects with the polarization control of the invisible illumination.

Radial diffraction of light in the radial momentum state space.

While linear or angular position and momentum can be linked by a continuous or discrete Fourier transform, there are some subtle problems in the analogous Fourier relationship between radial position and radial momentum in history. Here we exploit radial position and newly introduced radial momentum variables to report a radial version of light's diffraction. The mask with single or multiple radial slits confines the light to a radial transmission function. As a result, in the radial momentum state space, we can observe the diffraction sidebands generated on the transmitted light due to a transverse restriction of the radial range. Our experimental results clearly reveal the intriguing diffraction behaviors between radial position and radial momentum variables at the single-photon level, making them another candidate for fundamental tests of quantum mechanics and for a variety of quantum information applications.

Counting Crowds with Perspective Distortion Correction via Adaptive Learning.

The goal of crowd counting is to estimate the number of people in the image. Presently, use regression to count people number became a mainstream method. It is worth noting that, with the development of convolutional neural networks (CNN), methods that are based on CNN have become a research hotspot. It is a more interesting topic that how to locate the site of the person in the image than simply predicting the number of people in the image. The perspective transformation present is still a challenge, because perspective distortion will cause differences in the size of the crowd in the image. To devote perspective distortion and locate the site of the person more accuracy, we design a novel framework named Adaptive Learning Network (CAL). We use the VGG as the backbone. After each pooling layer is output, we collect the 1/2, 1/4, 1/8, and 1/16 features of the original image and combine them with the weights learned by an adaptive learning branch. The object of our adaptive learning branch is each image in the datasets. By combining the output features of different sizes of each image, the challenge of drastic changes in the size of the image crowd due to perspective transformation is reduced. We conducted experiments on four population counting data sets (i.e., ShanghaiTech Part A, ShanghaiTech Part B, UCF_CC_50 and UCF-QNRF), and the results show that our model has a good performance.

Realization of the Einstein-Podolsky-Rosen Paradox Using Radial Position and Radial Momentum Variables.

As is well known, angular position and orbital angular momentum (OAM) of photons are a conjugate pair of variables that have been extensively explored for quantum information science and technology. In contrast, the radial degrees of freedom remain relatively unexplored. Here we exploit the radial variables, i.e., radial position and radial momentum, to demonstrate Einstein-Podolsky-Rosen correlations between down-converted photons. In our experiment, we prepare various annular apertures to define the radial positions and use eigenmode projection to measure the radial momenta. The resulting correlations are found to violate the Heisenberg-like uncertainty principle for independent particles, thus manifesting the entangled feature in the radial structure of two-photon wave functions. Our work suggests that, in parallel with angular position and OAM, the radial position and radial momentum can offer a new platform for a fundamental test of quantum mechanics and for novel application of quantum information.

Exponential passivity of discrete-time switched neural networks with transmission delays via an event-triggered sliding mode control.

This paper investigates the exponential passivity of discrete-time switched neural networks (DSNNs) with transmission delays via an event-triggered sliding mode control (SMC). Firstly, a novel discrete-time switched SMC scheme is constructed on the basis of sliding mode control method and event-triggered mechanism. Next, a state observer with transmission delays is designed to estimate the system state. Moreover, some new weighted summation inequalities are further proposed to effectively evaluate the exponential passivity criteria for the closed-loop system. Finally, the effectiveness of theoretical results is showed through a simulative analysis on a multi-area power system.

Overexpressed CD39 mitigates sepsis­induced kidney epithelial cell injury via suppressing the activation of NLR family pyrin domain containing 3.

Unfettered inflammation is a leading cause of multiple organ failures in sepsis. The anti­inflammatory role of cluster of differentiation (CD)39 has been previously reported. The present study aimed to investigate the role of unfettered inflammation in sepsis­induced acute kidney injury (AKI). Lipopolysaccharide (LPS) was introduced to construct a sepsis mouse model. Kidney function and pathological changes in mice were measured at 12, 24 and 48 h. CD39 overexpression and inhibition vectors were transfected into renal tubular epithelial (HK­2) cells, followed by LPS treatment (10 µg/ml), and the cell viability changes at 24 h after treatment were assessed and the expression of NLR family pyrin domain containing 3 (NLRP3), cleaved caspase­1 and CD39 were determined by performing ELISAs. Cell apoptosis and reactive oxygen species (ROS) levels were determined by flow cytometry. It was found that after LPS administration, kidney injury was the most serious at 24 h in mice. CD39 overexpression could suppress the upregulation of pro­inflammatory cytokines induced by LPS treatment. In addition, the cell apoptosis and ROS level exhibited an obvious decrease, while cell viability increased. The NLRP3 expression and activity also showed a great inhibition in CD39­overexpressed cells. By contrast to CD39 overexpression, CD39 inhibition promoted the activation of the NLRP3 inflammasome. These data indicate the protective role of CD39 in LPS­induced renal tubular epithelial cell damage through inhibiting NLRP3 inflammasome activation and that CD39 might be a potential therapeutic target in sepsis­induced AKI.

Delay-dependent dynamical analysis of complex-valued memristive neural networks: Continuous-time and discrete-time cases.

This paper considers the delay-dependent stability of memristive complex-valued neural networks (MCVNNs). A novel linear mapping function is presented to transform the complex-valued system into the real-valued system. Under such mapping function, both continuous-time and discrete-time MCVNNs are analyzed in this paper. Firstly, when activation functions are continuous but not Lipschitz continuous, an extended matrix inequality is proved to ensure the stability of continuous-time MCVNNs. Furthermore, if activation functions are discontinuous, a discontinuous adaptive controller is designed to acquire its stability by applying Lyapunov-Krasovskii functionals. Secondly, compared with techniques in continuous-time MCVNNs, the Halanay-type inequality and comparison principle are firstly used to exploit the dynamical behaviors of discrete-time MCVNNs. Finally, the effectiveness of theoretical results is illustrated through numerical examples.

Convergence behavior of delayed discrete cellular neural network without periodic coefficients.

In this paper, we study convergence behaviors of delayed discrete cellular neural networks without periodic coefficients. Some sufficient conditions are derived to ensure all solutions of delayed discrete cellular neural network without periodic coefficients converge to a periodic function, by applying mathematical analysis techniques and the properties of inequalities. Finally, some examples showing the effectiveness of the provided criterion are given.