Search

Despite its simplicity and high accuracy, the real-time efficiency of OTSU segmentation is not high. In order to promote the real-timeliness of image segmentation, this paper introduces the fruitfly optimization algorithm (FOA) to OTSU segmentation, creating an FOA-OTSU segmentation algorithm. In the proposed algorithm, the optimal threshold for segmentation is searched for by the FOA. The classic Lena picture, Flower picture and Cameraman picture were used for simulation experiment, and the results were evaluated by signal-to-noise ratio (SNR) and peak signal-to-noise ratio (PSNR). The results show that the segmentation time is reduced by about 50.0% on the premise of the PSNR and SNR evaluation criteria and the segmentation effect basically unchanged. The simulations indicate that our method converges faster and consumes fewer time than traditional OTSU algorithm, without sacrificing the segmentation accuracy. The research provides a desirable tool with high real-time performance of rapid image segmentation.

Many vision-based applications rely on logistic regression for embedding classification within a probabilistic context, such as recognition in images and videos or identifying disease-specific image phenotypes from neuroimages. Logistic regression, however, often performs poorly when trained on data that is noisy , has irrelevant features , or when the samples are distributed across the classes in an imbalanced setting ; a common occurrence in visual recognition tasks. To deal with those issues, researchers generally rely on ad-hoc regularization techniques or model a subset of these issues. We instead propose a mathematically sound logistic regression model that selects a subset of (relevant) features and (informative and balanced) set of samples during the training process. The model does so by applying cardinality constraints (via $\ell _0$ l 0 -‘norm’ sparsity) on the features and samples. $\ell _0$ l 0 defines sparsity in mathematical settings but in practice has mostly been approximated (e.g., via $\ell _1$ l 1 or its variations) for computational simplicity. We prove that a local minimum to the non-convex optimization problems induced by cardinality constraints can be computed by combining block coordinate descent with penalty decomposition. On synthetic, image recognition, and neuroimaging datasets, we show that the accuracy of the method is higher than alternative methods and classifiers commonly used in the literature.

Generalized Zero-shot learning (GZSL) for object recognition is to solve the problem of recognizing samples in both seen and unseen classes, while only having seen classes in training. As labeling abundant examples for all kinds of classes in realistic scenarios is costly and impractical, GZSL has become a novel and hot field of research in recent years. Most previous methods tried to learn a fixed one-directional mapping, either from visual to semantic features, or from semantic to visual features. However, recently, cross mapping between visual and semantic features has achieved good results and many methods come up based on this idea. In this paper, we propose a novel model. It is Multi-channel Gaussian Mixture VAE(MCGM-VAE), which introduces Gaussian mixture model to our multi-modal VAE with multiple channels. These channels are of different weight coefficients following with channel-weight layers, so as to produce a Gaussian mixture distribution. Then the latent space could be generated from it. We evaluate our method on several benchmark databases, i.e. CUB, SUN, AWA1, AWA2, aPY and prove our approach outperforms state-of-the-art methods. Through the experimental data analysis, the impact of some hyperparameters on the experimental performance is further analyzed.

We experimentally demonstrate an integrated strictly non-blocking silicon 4 × 4 optical switch chip that can be operated in both thermo-optic (TO) and electro-optic (EO) switching modes. It is based on the double-layer network (DLN) architecture and consists of twelve 2 × 2 Mach-Zehnder interferometer (MZI) switch elements. TO phase shifters based on TiN microheaters and EO phase shifters based on p-i-n diodes are embedded in both waveguide arms of the MZI elements. The power consumption for TO and EO switching is 34 mW/π and 7 mW/π, respectively. The on-chip insertion losses are 1.74 ± 0.59 dB and 3.79 dB ± 1.32 dB for TO and EO switching, respectively. Due to the merits of the DLN architecture and the optimized performance of the switch elements, the chip possesses low crosstalk of -29.1 dB and -19.4 dB for TO and EO switching, respectively. Quadrature phase-shift keying (QPSK) optical signals with a data rate of 64 Gb/s are transmitted through the switch with no observable deteriorations. Such an optical switch is a promising candidate for both optical circuit switching and optical packet switching for a variety of applications.

Stickers are used by social users to express online chat. The retrieval, matching, and generation of stickers have significant research and application value. However, the current stickers are mainly obtained through “semantic-visual” cross-modal search matching in a pre-defined sticker library. The retrieved images from the pre-defined sticker library are fixed and limited, and often do not conform to the contextual semantics of chatting. This work focuses on synthesizing non-existent stickers instead of retrieving images. The Object-Orient Semantic-to-Visual Generation Model(OSVGAN) is proposed to instantly synthesize the corresponding sticker according to the user's semantic description. First, the intrinsic semantic relationship of each object in the sticker is established by an Object-Orient Meta Model. Then a Semantic-to-Visual model is used to synthesize corresponding images according to each object's text feature, and a Fusion Model fuses each object image to form a complete image. Finally, to enhance the semantic consistency of the generated images, the text-image alignment module(T2I) is applied to the Semantic-to-Visual model and the Fusion Model. OSVGAN can effectively reduce the demand for training samples. On the sticker dataset, our method has achieved good semantic consistency, and is comparable to the existing methods in terms of synthetic image quality.

Bioink, a key element of three-dimensional (3D) bioprinting, is frequently engineered to achieve improved printing performance. Viscoelasticity related to rheological properties is correlative of the printability of bioink for extrusion bioprinting, which affects the complexity of printing 3D structures. This article shows the use of hydroxyethyl cellulose (HEC) as a rheological additive for engineering bioink to improve the printability without reducing the biocompatibility. Different concentrations of HEC were added to four types of bioink, namely, reagent-crosslinked, temperature-dependent phase change, ultraviolet-polymerized, and composite hydrogel bioinks, to investigate the effect on the viscoelasticity properties, print fidelity, and other printed scaffold properties. The results indicate that HEC is able to increase the rheological properties by 100 times to stabilize complex structures and improve the printing fidelity to narrow the gap between the design value and theoretical value, even converting nonviscous ink into directly printable ink, as well as tune the swelling ratio for better molecular permeability. The degradation of bioink can also be tuned by the addition of HEC. Moreover, this bioink is biocompatible for cell lines and primary cells. HEC is expected to be widely used in 3D extrusion-based bioprinting.

Psoriasis is an immune-mediated chronic inflammatory skin disease primarily mediated by the activation of interleukin (IL)-17-producing T cells. Traditional Chinese Medicine (TCM) represents one of the most effective complementary and alternative medicine (CAM) agents for psoriasis, which provides treasured sources for the development of anti-psoriasis medications. Xiao-Yin-Fang (XYF) is an empirically developed TCM formula that has been used to treat psoriasis patients in Shanghai Changhai Hospital for over three decades. Imiquimod (IMQ)-induced psoriasis-like dermatitis mouse model was utilized to investigate the therapeutic effects of XYF by the assessment of disease severity and skin thickness. Flow cytometric assay was performed to explore the influence of XYF on skin-related immunocytes, primarily T cells. And, RNA sequencing analysis was employed to determine the alternation in gene expression upon XYF therapy. We discovered that XYF alleviated psoriasis-like skin inflammation mainly through suppressing dermal and draining lymph-node IL-17-producing γδT (γδT17) cell polarization. Moreover, XYF therapy ameliorated the relapse of psoriasis-like dermatitis and prohibited dermal γδT cell reactivation. Transcriptional analysis suggested that XYF might regulate various inflammatory signaling pathways and metabolic processes. In conclusion, our results clarified the therapeutic efficacy and inner mechanism of XYF therapy in psoriasis, which might promote its clinical application in psoriasis patients and facilitate the development of novel anti-psoriasis drugs based on the bioactive components of XYF.

Real-time and accurate vehicle tracking by Cameras and Surveillance can provide strong support for the acquisition and application of important traffic parameters, which is the basis of the traffic condition evaluation and the reasonable traffic command and dispatch. To deal with difficult problems of vehicle tracking research in a complex environments, such as occlusion, sudden illumination change, similar target interference and real-time tracking, measures are taken as follows. Firstly, the existing color local entropy particle filter tracking method is improved. The symmetry of information entropy is used to overcome the tracking failure caused by large-area occlusion. Secondly, the SIFT feature tracking method is improved to enhance real-time performance and robustness. Thirdly, two tracking methods were combined according to their characteristics, aiming at effectively improving the quasi-determination and real-time performance of vehicle tracking. Fourthly, Kalman filter was used to predict the motion state of vehicles. According to the SIFT characteristics and license plate information of vehicles, the exact position of the lost target vehicles is quickly located. It has been verified by experiments that our method has effectively improved the accuracy and real-time performance of vehicle tracking in complex situations.

Although deep convolutional neural network (CNN) performs well in many computer vision tasks, its classification mechanism is very vulnerable when it is exposed to the perturbation of adversarial attacks. In this paper, we proposed a new algorithm to generate the substitute model of black-box CNN models by using knowledge distillation. The proposed algorithm distills multiple CNN teacher models to a compact student model as the substitution of other black-box CNN models to be attacked. The black-box adversarial samples can be consequently generated on this substitute model by using various white-box attacking methods. According to our experiments on ResNet18 and DenseNet121, our algorithm boosts the attacking success rate (ASR) by 20% by training the substitute model based on knowledge distillation.

The objective of digital broadcasting has evolved from providing a plain video service to offering a realistic visual experience. Technologies such as 3DTV and UHDTV have been suggested to achieve this new objective by providing an immersive and stereoscopic visual experience. However, owing to the high bandwidth requirements of such services, the broadcasting industry has faced a challenge to find a new transport mechanism for overcoming the bandwidth limitation. The standardization organizations, the Advanced Television Systems Committee, Digital Video Broadcasting, and Telecommunications Technology Association, have been working on the integration of broadcasting and a broadband network (IP) to resolve the bandwidth issue of realistic video services. This paper introduces a frame-level timeline synchronization and transport system target decoder model for providing a stable 3DTV service over a hybrid network. The experimental results indicate that the proposed technologies can be successfully adopted as a reference model in a broadcast-broadband hybrid 3DTV service and other IP-associated hybrid broadcasting services.

The concept of minimodeling is defined as a kind of focal bone formation that features the absence of preceding bone absorption by osteoclasts. In the process of minimodeling, osteogenetic and osteoclastic activities are decoupled. Linear boundary between old bone and new bone can be discovered. Frost et al. presumed that minimodeling in trabeculae can continue throughout life. The concept of minimodeling is not new, however its function and nature are still imperfectly understood. Our review will focus on minimodeling-based formation in 5 aspects below: 1. compare the minimodeling with remodeling regarding 10 items; 2. describe the histological observation to characterize remodeling and minimodeling formation sites; 3. present evidence of bone anabolic agents which start bone minimodeling; 4. discuss the mechanism and target cells involved in bone minimodeling; 5. interpretate the increased osteogenic potential and trabecular connectivity with minimodeling-based formation; 6. Depict modeling patterns of the bone implant interface and peri-implant osteogenesis in osteoporosis. This review provides an in-depth knowledge regarding minimodeling comprehensively and its potential contributions in skeletal homeostasis and application prospect in improving the success rate of dental implant in osteoporosis.

We demonstrate a polarization-insensitive silicon 4×4 optical switch based on Mach–Zehnder interferometer (MZI) switch elements. On-chip polarization controllers are integrated before the switch fabric to automatically adjust an arbitrary input polarization to the transverse electric mode. The 4×4 switch fabric is based on a dilated double-layer network architecture to completely cancel the first-order crosstalk. Thermo-optic phase shifters are integrated in the MZI switch elements and the polarization controllers for adjustment of the switching state and polarization, respectively. We develop a polarization control algorithm based on a gradient descent method for automated polarization control. The polarization recovery time is less than 4 ms, and the measured polarization-dependent loss is ∼2 dB. The scheme provides a new solution for realizing polarization-insensitive silicon optical switches.

Individual tree segmentation is essential for many applications in city management and urban ecology. Light Detection and Ranging (LiDAR) system acquires accurate point clouds in a fast and environmentally-friendly manner, which enables single tree detection. However, the large number of object categories and occlusion from nearby objects in complex environment pose great challenges in urban tree inventory, resulting in omission or commission errors. Therefore, this paper addresses these challenges and increases the accuracy of individual tree segmentation by proposing an automated method for instance recognition urban roadside trees. The proposed algorithm was implemented of unmanned aerial vehicles laser scanning (UAV-LS) data. First, an improved filtering algorithm was developed to identify ground and non-ground points. Second, we extracted tree-like objects via labeling on non-ground points using a deep learning model with a few smaller modifications. Unlike only concentrating on the global features in previous method, the proposed method revises a pointwise semantic learning network to capture both the global and local information at multiple scales, significantly avoiding the information loss in local neighborhoods and reducing useless convolutional computations. Afterwards, the semantic representation is fed into a graph-structured optimization model, which obtains globally optimal classification results by constructing a weighted indirect graph and solving the optimization problem with graph-cuts. The segmented tree points were extracted and consolidated through a series of operations, and they were finally recognized by combining graph embedding learning with a structure-aware loss function and a supervoxel-based normalized cut segmentation method. Experimental results on two public datasets demonstrated that our framework achieved better performance in terms of classification accuracy and recognition ratio of tree.

Robotic assisted technologies benefit surgeon for less workload via dexterous manipulation in minimally invasive surgeries (MIS). Redundant manipulator can provide sufficient dexterity and is potential to use in a confined and narrow workspace. We had designed a master-slave paediatric surgical assist robot, which consists of two symmetric slave arms with nine DOFs each for dexterous manipulation. This paper presents the kinematics computation of the redundant manipulator based on its geometrical structure, and then, a shape optimization algorithm is constructed to calculate the posture configuration along the redundant manipulator, subsequently, a collision avoidance algorithm is constructed to perform motion planning in a narrow workspace. The experiments were executed to validate the correctness of the derived inverse kinematics algorithm and the results demonstrated that the redundant manipulator is capable of performing dexterous manipulation with a positioning accuracy within 1.5 mm in a narrow surgical workspace.

The present study is firstly designed to identify the relationship between serum omentin-1 concentration, body fat mass and bone mineral density in healthy Chinese male adults in Changsha city. A total of 219 (20–80 years old) healthy subjects were enrolled in this cross-sectional study. Serum omentin-1, adiponectin, leptin, resistin and bone turn over biochemical markers were measured with enzyme-linked immunosorbent assay. Bone mineral density (BMD) and fat body composition were determined using dual-energy-X-ray absorptiometry. Serum omentin-1 levels in the overweight subjects were significantly lower than those of the subjects with normal weight (p

To achieve the flexible and efficient transmission of files in the video on demand system, this paper proposed a three-tier fragmentation strategy which provided a unfiled method of fragmentation for video of HTTP Live Streaming and large file. In order to prove the rationality of the three-tier fragmentation strategy, we compared it with the two layers fragmentation strategy and the fragmentation strategy of equal time. And the comparisons proved the superiority of three-tier fragmentation strategy.

Objective This study aims to identify effective gene networks and biomarkers to predict response and prognosis for HER2-negative breast cancer patients who received sequential taxane-anthracycline neoadjuvant chemotherapy. Materials and Methods Transcriptome data of training dataset including 310 HER2-negative breast cancer who received taxane-anthracycline treatment and an independent validation set with 198 samples were analyzed by weighted gene co-expression network analysis (WGCNA) approach in R language. Gene ontology (GO) terms and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways analysis were performed for the selected genes. Module-clinical trait relationships were analyzed to explore the genes and pathways that associated with clinicopathological parameters. Log-rank tests and COX regression were used to identify the prognosis-related genes. Results We found a significant correlation of an expression module with distant relapse–free survival (HR = 0.213, 95% CI [0.131–0.347], P = 4.80E−9). This blue module contained genes enriched in biological process of hormone levels regulation, reproductive system, response to estradiol, cell growth and mammary gland development as well as pathways including estrogen, apelin, cAMP, the PPAR signaling pathway and fatty acid metabolism. From this module, we further screened and validated six hub genes (CA12, FOXA1, MLPH, XBP1, GATA3 and MAGED2), the expression of which were significantly associated with both better chemotherapeutic response and favorable survival for BC patients. Conclusion We used WGCNA approach to reveal a gene network that regulate HER2-negative breast cancer treatment with taxane-anthracycline neoadjuvant chemotherapy, which enriched in pathways of estrogen signaling, apelin signaling, cAMP signaling, the PPAR signaling pathway and fatty acid metabolism. In addition, genes of CA12, FOXA1, MLPH, XBP1, GATA3 and MAGED2 might serve as novel biomarkers predicting chemotherapeutic response and prognosis for HER2-negative breast cancer.

Metastatic disease is the major cause of death from cancer, and immunotherapy and chemotherapy have had limited success in reversing its progression. Researchers have suggested that inflammatory factors in the tumor environment can promote cancer invasion and metastasis, stimulating cancer progression. Thus, novel strategies that target cytokines and modulate the tumor microenvironment may emerge as important approaches for treating metastatic breast cancer. Specific neutralization of pathogenic TNF signaling using a TNFα antibody has gained increasing attention. Considering this, a selective human TNFα neutralized antibody was generated based on nanobody technology. A TNFα-specific nanobody was produced in Pichia pastoris with a molecular mass of 15 kDa and affinity constant of 2.05 nM. In the proliferation experiment, the TNFα nanobody could inhibit the proliferation of the breast cancer cell line MCF-7 induced by hTNFα in a dose-dependent manner. In the microinvasion model, the TNFα nanobody could inhibit the migration of the breast cancer cell lines MCF-7, MDA-MB-231 and the invasiveness of MDA-MB-231 induced by hTNFα in a dose-dependent manner. Drug administration of the combination of paclitaxel with the TNFα nanobody in vivo significantly enhanced the efficacy against 4T-1 breast tumor proliferation and lung metastasis; meanwhile, E-cadherin tumor epithelial marker expression was upregulated, supporting the anti-tumor therapeutic relevance of paclitaxel and the TNFα nanobody on EMT. This study highlights the importance of neutralizing low TNFα levels in the tumor microenvironment to sensitize the chemotherapeutic response, which has attractive potential for clinical applications.

Nowadays, the users of IPTV require better user experience and the media providers are interested in finding the key factors which will influence the Quality of Experience (QoE) and the way to predict the QoE. In this paper, we discuss the relationship between the status of IPTV set-top box and user's QoE. We first clean the dataset and conduct some statistical analysis in dataset. Then, we select the important features in key performance indicator (KPI) dataset by the feature selection technology. In addition, we compare the decision tree model and Adaboost model in KPI -- Cuser's complaint dataset which is a classical imbalanced dataset. Extensive experimental results demonstrate that in imbalanced KPI-user's complaint dataset Adaboost model performs better than the decision tree in terms of predicting the user's complaint, and importantly we design a cost coefficient for Adaboost model to achieve a higher accuracy.

An optical fiber curvature sensor based on photonic crystal fiber (PCF) modal interferometer is demonstrated. The sensor is fabricated by a length of 5 cm PCF which is spliced between two single mode fibers (SMF). The result shows that this sensor is sensitive to curvature with a sensitivity coefficient of 4.451 nm m−1 in the range of 0–2.14 m−1, and the resolution of the curvature measurement is about 4.4 × 10−3 m−1. The temperature sensitivity of the sensor is only 7.78 pm/°C, so it is more suitable for curvature measuring.

A modal interferometer based on a 1.2 mm hollow-core photonic crystal fiber (HC-PCF) for strain measurement is presented. The modal interferometer is fabricated by splicing the two ends of the HC-PCF to single-mode fibers (SMF). The air-holes in the core and cladding of the HC-PCF are fully collapsed in the splicing process, and the length of each collapsed region is about 300 μm. The transmission spectra with different strain levels are measured. The sensitivity of 1.83 pm/μɛ is achieved in the range of 0–1260 μɛ, and the resolution for strain measurement was about 11 μɛ for wavelength demodulation.

With the rapid development of modern communication technology, all kinds of radio technologies emerge endlessly. As different wireless networks share the same spectrum, coexistence of the heterogeneous networks becomes a challenging problem, especially the coexistence between ZigBee and WiFi. Much research has been done to solve this problem, and some coexistence methods have been proposed, such as the spectrum allocation, power control and so on. However, the proposed methods are either too complex to implement or need additional hardware cost. This paper proposes an easy and feasible method called AHA for ZigBee. The method modifies the current FEC coding method and additionally colligates energy sensing and adaptive retransmission. By implementing the method on ZigBee using USRP N210 devices, it is demonstrated that AHA effectively improves the communication performance of ZigBee under WiFi interference.

This study was devised to investigate the genetic effect modification of the BsmI polymorphism associated with the susceptibility to breast cancer. Case-control studies of the BsmI polymorphism and breast cancer were searched. A total of 17 eligible publications were included in our final analysis. Pooled ORs and 95 % CIs were obtained by means of fixed effects model. The general and stratified analyses according to ethnicity showed that the association between the BsmI polymorphism and the risk of breast cancer was not statistically significant. However, the subgroup of the hospital-based studies was found to confer protection against the disease (ORBBvs.bb = 0.83, 95 % CI = 0.71-0.97, P h = 0.571; OR BBvs.Bb+bb = 0.86, 95 % CI = 0.74-1.00, P h = 0.903; OR allele B vs. allele b = 0.92, 95 % CI = 0.86-0.99, P h = 0.337). Our results suggested that the presence of the BsmI polymorphism may contribute to the susceptibility of breast cancer. It is necessary that future large-scale studies should be conducted to further confirm the association between the BsmI polymorphism and breast cancer risk.

Geosynchronous Synthetic Aperture Radar (GEO SAR) is the satellite running on the 36,500-Km altitude orbit, and its period is almost 1-day. As we all know, if there is no effect of perturbation force, the orbit of the GEO SAR will coincident exactly per period, and there will be no baseline to achieve interferometry. However in practice, there are varieties of perturbation force in the space, so the orbit in different periods is no longer coincident, which makes it possible to achieve interferometry. In this paper, we study two typical GEO SAR orbits, and analyze the nadir-point trajectory excursion due to orbital perturbation using the software STK (Satellite Tool Kits, AGI). Based on the interferometry theory, we define the repeat-pass baseline of the GEO SAR, and obtain the height measurement accuracy of different kinds of orbits. At last, we verify the feasibility of GEO SAR interferometry by simulation.

Mutiwavelength lasers are of great interest as their wide applications. In this paper, a novel multiwavelength pulse Brillouin Erbium-doped fiber laser was demonstrated and its principle was given. An electro-optic modulator (EOM) was used to modulate continue wave into pulse light. The results show that, the peak values are unstable, about ±3dB, but the wavelength interval was stable, about 0.088nm. And the unstable factor was discussed. Also the influence of the Brillouin pump power on output optical spectrum was discussed in experiment. Both the wavelength number and the power of the output light increased with the increasing of pulse Brillouin pump power.

The theory of fiber Brillouin amplifier is investigated by numerical solution combinating interpolation method and Runge-Kutta method. Meanwhile, the most complete characterization and comparison of FBAs for pump power, initial signal power and fiber length is obtained. Through the analysis and comparison, the results show that the amplification efficiency can reach 90% based on the SBS and the most of the power transfer occurs within the first 20% of the fiber length. The output signal power is linear increased with increasing of the initial pump and initial signal, respectively. The high-gain Brillouin amplifier can be obtained when then fiber length is about 26km.

The influence of forward Raman amplification on Brillouin scattering signal of pulse light was investigated in experiments, including the spontaneous Brillouin scattering and the stimulated Brillouin scattering (SBS). The results of experiments show that the spontaneous Brillouin scattering is amplified by fiber Raman amplifier and fiber Brillouin amplifier, consequently the gain is high. For the amplification of SBS signal, it is only amplified by Raman amplifier in the fiber. As the Raman pump power reaches 600mW, it causes multi-order Brillouin scattering, and then the gain of the first order Stokes SBS starts to saturate and decline.

The influence of the fiber length and input power on pulse light stimulated Brillouin scattering effect was investigated. Laser with the wavelength 1550nm emitted a continuous optical input electro-optic modulator, and then is modulated into pulsed light. The pulsed signal is provided by the signal generator, pulse width 200ns, repetition frequency 2 KHz. The results show that the threshold of pulse light stimulated Brillouin scattering is 3.1dBm with the 25 Km single-mode optical fiber, and the light intensity of stimulated Brillouin scattering starts to grow slowly when the pump power is beyond 8.1dBm. It can be seen that Brillouin Stokes power increases with the growth of fiber length, but when the fiber is longer than 31km, the Brillouin Stokes power reaches saturation.