Search

Video-based person re-identification (Re-ID) aims at retrieving the person through the video sequences across non-overlapping cameras. Some characteristics of pedestrians are not consecutive across frames due to the variations of viewpoints, postures, and occlusions over time. However, existing methods ignore such data peculiarity and the networks tend to only learn those salient consecutive characteristics among frames in video sequences. As a result, the learned representations fail to cover all the characteristics of pedestrians, thus lacking integrity and discrimination. To tackle this problem, we present a novel deep architecture termed Hierarchical Mining Network (HMN), which mines as many pedestrians’ characteristics by referring to the temporal and intra-class knowledge. It consists of a novel Attentive Temporal Module (ATM) and a Dynamic Supervising Branch (DSB), with a Balancing Triplet Loss (BTL) assisting the training. The proposed ATM, with pedestrian perceiving capacity, is capable of evaluating each activation of features through temporal analysis, so that the temporally scattered characteristics of pedestrians can be better aggregated and the contaminated ones can be eliminated. Then, the DSB along with the BTL further enhances the integrity of representations by multiple supervision. Specifically, the DSB perceives the diversities of intra-class samples in each mini-batch and generates targeted supervising signals for them, in which process the BTL guarantees the signals with smaller intra-class variations and larger inter-class variations. Comprehensive experiments on two video-based datasets, i.e., MARS, and DukeMTMC-VideoReID, demonstrate the contribution of each component and the superiority of the proposed HMN over the state-of-the-arts. Benchmarking our model on three popular image-based datasets, i.e., Market1501, DukeMTMC-Reid, and MSMT17 additionally verifies the promising generalizability of the proposed DSB and BTL.

The ability to simulate vegetation dynamics and their feedback with nutrient cycling to affect ecosystem productivity underpins our prediction of the land carbon sink under climate change. Predictive models are now capable of simulating complex ecosystem processes, including the recent advancement in simulating vegetation dynamics and ecosystem phosphorus cycling, but there is a general lack of empirical evidence to form a systematic evaluation of the model predictions, especially how functional diversity affect ecosystem nutrient cycling and its consequence for productivity. Here, we developed a dataset based on 9 permanent plots (20 x 20 m) along an elevation gradient (300 – 1200m a.s.l.) in a subtropical forested mountain in eastern China. We measured vegetation growth, estimated forest structure and species composition, and compiled ecosystem-scale carbon (C), nitrogen (N) and phosphorus (P) budgets based on concentration, pool and flux data collected from dominant canopy trees, understorey herbaceous plants, and soil organic and inorganic components in these forested plots. Our aims are three-fold: 1) to understand how C, N and P are distributed along the plant-microbe-soil continuum; 2) to disentangle how different growth and nutrient use strategies of plant and soil microbes affect ecosystem productivity and regulate the rate nutrient cycling; and 3) to benchmark predictive models in simulating ecosystem vegetation dynamics and their interaction with C, N, and P cycle processes. Our research will contribute towards better understanding of the functional diversity and productivity relationship, and will contribute towards an improved predictive capacity to simulate vegetation dynamics and the land carbon sink under climate change.

Carbon nanomaterials, such as carbon nanoonions (CNOs), possess promising applications in various fields. There are urgent demands to synthesize carbon nanomaterials from a green and renewable carbon source. In this study, solid CNOs with relatively uniform size distribution (with diameters of about 30–50 nm), abundant structure defects and oxygen-containing surface functional groups (such as –OH and –COOH) are developed from co-pyrolysis of lignin (LG) and polyethylene (PE) in the presence of Ni-based catalysts. The type of catalyst, the concentration of catalyst and catalytic co-pyrolysis temperature play important roles in the morphologies and properties of CNOs as confirmed by TEM and SEM. Furthermore, the produced CNOs can act as a low-cost and highly-efficient adsorbent to remove Cu(II) from aqueous solution according to a homogeneous monolayer, chemical action-dominated, endothermic and spontaneous process. The theoretical maximum adsorption capacity of CNOs calculated from the Langmuir model is 100.00 mg g−1. Surface deposition, complexation, π electron–cation interaction and electrostatic interaction are responsible for the adsorption of Cu(II) using the prepared CNOs.

Arrhythmia is one of the common cardiovascular diseases. Nowadays, many methods identify arrhythmias from electrocardiograms (ECGs) by computer-aided systems. However, computer-aided systems could not identify arrhythmias effectively due to various the morphological change of abnormal ECG data. This paper proposes a deep method to classify ECG samples. Firstly, ECG features are extracted through continuous wavelet transform. Then, our method realizes the arrhythmia classification based on the new lightweight context transform blocks. The block is proposed by improving the linear content transform block by squeeze-and-excitation network and linear transformation. Finally, the proposed method is validated on the MIT-BIH arrhythmia database. The experimental results show that the proposed method can achieve a high accuracy on arrhythmia classification.

Correct and continuous weight measurement of a burn patient is important for fluid resuscitation. Generally, a body weight bed uses four sensors to obtain signals. However, owing to signals dynamic characteristics, the air fluidized therapy bed has a large error in this mode, which is difficult to reduce. In this study, we optimized the sensor configuration scheme and developed a set of signal processing and methods to evaluate noise reduction. Additionally, the wavelet neural network was used to improve the weighing accuracy. Furthermore, medical tools are located on different body parts of a patient to acquire weight characteristics to assist with the rehydration of patient. The average error of 183.33g is acceptable from a medical perspective. An average error of 250.21g in dynamic weight after using wavelet de-noising and wavelet neural network methods was obtained.

Soil organic carbon (SOC) in deeper soils may be more protected by mineral particles than SOC in shallower soils; thus, the vertical SOC distribution is closely related to the overall SOC stability. This study aimed to verify the relationship between the solubility temperature dependence of water extractable SOC fractions and their vertical distribution characteristics. Soil samples collected from each soil horizon in natural dragon spruce (Picea asperata) and red birch (Betula albosinensis) forests were analyzed. Cold water-extracted organic carbon (CWEOC) and hot water-extracted organic carbon (HWEOC) were extracted at 20 °C and 80 °C, respectively. The sum of CWEOC and HWEOC was considered the water-extracted organic carbon (WEOC) content. The carbohydrate-C (Car-C) and phenolic-C (Phe-C) contents extracted by hot water were also determined. The CWEOC, HWEOC, Car-C, and Phe-C contents varied significantly (P

To retrospectively evaluate whether repositioning the bone window leads to a better outcome of three-dimensional sinus augmentation in lateral sinus floor elevation (LSFE) with simultaneous implant placement.34 patients with a total of 40 implants (14: test group, 26: control group) receiving LSFE with simultaneous implant placement were included in this retrospective research. CBCT images were taken before surgery, immediately and 6 months after surgery. The two-dimensional augmentation parameters, including apical bone height (ABH), endo-sinus bone gain (ESBG), and palatal/buccal bone height (PBH/BBH), and three-dimensional parameters, including augmentation volume (AV) and palatal/buccal augmentation volume (PAV/BAV), were measured. The lateral defect length (LDL) and lateral window length (LWL) were also measured to evaluate the lateral antrostomy recovery.At the 6-month follow-up, the reduction rates at ABH, ESBG, and BBH of the test group (ABH: 10.41% ± 30.30%, ESBG: 2.55% ± 8.91%, BBH: 2.50% ± 8.65%) were significantly lower than those of the control group (ABH: 25.10% ± 22.02%, ESBG: 11.47% ± 9.79%, BBH: 7.10% ± 5.37%; p .05). In addition, the test group showed better three-dimensional augmentation stability on the buccal side (BAV reduction: 15.51% ± 10.86% vs. 27.15% ± 12.61%; p .05). Moreover, the LDL/LWL ratio of the test group was significantly lower than that of the control group (p .05).Within the limitations of this study, repositioning of the bone window in LSFE with simultaneous implant placement could contribute to endo-sinus augmentation stability on the buccal side at the 6-month follow-up. Moreover, it would also facilitate recovery of the lateral antrostomy defect.

Objective. With the improvement of living standards, heart disease has become one of the common diseases that threaten human health. Electrocardiography (ECG) is an effective way of diagnosing cardiovascular diseases. With the rapid growth of ECG examinations and the shortage of cardiologists, accurate and automatic arrhythmias classification has become a research hotspot. The main purpose of this paper is to improve accuracy in detecting abnormal ECG patterns. Approach. A hybrid 1D Resnet-GRU method, consisting of the Resnet and gated recurrent unit (GRU) modules, is proposed to implement classification of arrhythmias from 12-lead ECG recordings. In addition, the focal Loss function is used to solve the problem of unbalanced datasets. Based on the proposed 1D Resnet-GRU model, we use class-discriminative visualization to improve interpretability and transparency as an additional step. In this paper, the Grad-CAM++ mechanism has been employed to the trained network model and generate thermal images superimposed on raw signals to explore underlying explanations of various ECG segments. Main results. The experimental results show that the proposed method can achieve a high score of 0.821 (F1-score) in classifying 9 kinds of arrythmias, and Grad-CAM++ not only provides insight into the predictive power of the model, but is also consistent with the diagnostic approach of the arrhythmia classification. Significance. The proposed method can effectively select and integrate ECG features to achieve the goal of end-to-end arrhythmia classification by using 12-lead ECG signals, which can serve a promising and useful way for automatic arrhythmia classification, and can provide an explainable deep leaning model for clinical diagnosis.

Dendrobium officinale flos (DOF) is the flower of Dendrobium officinale Kimura et Migo, which is usually regarded as a by-product of Dendrobii Offcinalis Caulis. Based on its use as an alternative medicine, we evaluated the antidepressant-like effect of DOF extracts on chronic, unpredictable, mild stress-induced, depression-like behaviour in mice and tested the effects of DOF on the regulation of neurotrophic factors in mouse astrocyte primary cultures and PC12 cell lines. Oral treatment with DOF ethanol extract (DOF-E) could alleviate depression-like behaviours in stress-exposed mice, as evidenced by increased sucrose consumption and decreased immobile time in a forced swim test. In the hippocampus, DOF extracts increased the expression of NGF and BDNF, both at the transcriptional and protein levels. In astrocytes, DOF-E increased the expression of NGF and BDNF via a cAMP-dependent mechanism and regulated plasminogen and matrix metallopeptidase 9 (MMP-9), which are related to the metabolic regulation of neurotrophic factors. In PC12 cells, DOF-E induced the expression of neurofilaments and potentiated the induction of neurite outgrowth upon treatment with a low dose of NGF. Based on these findings, DOF might be used as a supplement for antidepressant therapy in patients with depression.

In the present study, the coked catalysts derived from catalytic reforming of the pyrolysis volatiles of polyethylene (PE), lignin (LG) and their mixture were developed as low-cost and environmentally-friendly carbon materials-containing composites to remove heavy metal ions from aqueous solution. The composites were thoroughly characterized by SEM, TEM, XRD, TGA and FT-IR and then their adsorption capability towards Pb(ii) was investigated. It is found that curved cone-shape carbon nanotubes (CNTs) with abundant structural defects and O-containing surface functional groups, such as C-O, CO and -OH, can be obtained from the catalytic reforming of the mixture of PE and LG. The CNT-containing catalyst composite presents a superior adsorption capability towards Pb(ii) when it is employed in Pb(ii) removal. Adsorption isotherm and adsorption kinetics studies show that the adsorption process can be well simulated by the Langmuir isotherm and pseudo-second-order model, demonstrating that the adsorption is subjected to a homogeneous and chemical process. The calculated maximum adsorption capacity is as high as 146.08 mg g-1, which is much higher than most of the adsorbents reported. Moreover, thermodynamic analysis reveals that the adsorption is spontaneous and endothermic. Accordingly, the used catalyst from the catalytic reforming can be developed as a low-cost and highly-efficient adsorbent.

Lithium–sulfur batteries are regarded as the most promising electrochemical energy storage device as a result of their satisfactory high specific capacity and high energy density values. However, the inferior conversion efficiency of lithium polysulfides (LiPSs) in essence leads to fast capacity decay, especially at a large current density during the charge–discharge process. Herein, we have successfully demonstrated that the conversion efficiency of LiPSs was remarkably promoted by employing nickel nanoparticles incorporated with nitrogen-doped multi-channel carbon nanofibers (Ni-NMCCF) as a multifunctional modification layer of the separator. The high polarity of the nitrogen atoms and the Lewis acid characteristics of the nickel atoms collectively create strong anchoring activity towards LiPSs, therefore leading to the severe inhibition of the occurrence of the shuttle effect. The electrocatalytic activity of the nickel nanoparticles greatly accelerates the reversible conversion efficiency of LiPSs, ensuring high capacity retention during long-term cycling. Thanks to these advantages, Ni-NMCCF delivered a superb rate performance of 1020 mA h g−1 at 1C and maintained 671 mA h g−1 even after 500 cycles, which is equivalent to a capacity decay of only 0.07% per cycle.

Sandy soils in tropical areas are more vulnerable to potential toxic elements as a result of their low nutrition. The composite addition of biochar and phosphate material is considered a promising method of immobilizing toxic metals in sandy soils, but the synergistic effects of this process still need to be further explored, especially in typical tropical vegetables. In this study, a pot experiment was conducted to evaluate the agronomic and toxic metal-immobilization effects of single amendments (phosphate rock, activated phosphate rock, and biochar) and combined amendments, including biochar mixed with phosphate rock (BCPR) and biochar mixed with activated phosphate rock (BCAPR), on vegetables grown in tropical sandy soil. Among these amendments, the composite amendment BCAPR was the most effective for increasing Ca, Mg, and P uptake based on water spinach (Ipomoea aquatica L.) and pepper (Capsicum annuum L.), showing increased ratios of 22.5%, 146.0%, and 136.0%, respectively. The SEM-EDS and FTIR analysis verified that the activation process induced by humic acid resulted in the complexation and chelation of the elements P, Ca, and Mg into bioavailable forms. Furthermore, the retention of available nutrition elements was enhanced due to the strong adsorption capacity of the biochar. In terms of cadmium (Cd) and lead (Pb) passivation, the formation of insoluble mineral precipitates reduced the mobility of these metals within the BCAPR treatments, with the maximum level of extractable Cd (86.6%) and Pb (39.2%) reduction being observed in the tropical sandy soil. These results explore the use of sustainable novel cost-effective and highly efficient bi-functional mineral-based soil amendments for metal passivation and plant protection.

Li–S batteries have been attracting increasing interest owing to their remarkable advantages of low cost, high theoretical capacity and high theoretical energy density. Nevertheless, the severe “shuttle effects” of lithium polysulfides have markedly limited the performance of the cells and further hindered their commercial applications. Herein, a novel scheme combining a transformation strategy with ammonia treatment was developed to fabricate ZIF-8-derived nitrogen-doped hierarchically porous carbon (NHPC/NH3). When NHPC/NH3 was used as a host of sulfur, the obtained S@NHPC/NH3 cathode for Li–S cells presented an initial specific capacity of 1654 mA h g−1 and an outstanding cycling stability with only 0.27% attenuation per cycle from the 30th cycle to 130th cycle. Together with the theoretical calculation, it was concluded that such excellent electrochemical performances should be attributed to the suppressed “shuttle effect” via both physical and chemical adsorption of lithium polysulfides in the optimized microporous structures with effective nitrogen doping sites as well as the improved kinetics owing to the abundant meso/macroporous structures.

Algal blooms have posed great threats to livestocks and human health. Although flocculation is effective, its efficiency may hinder the direct application for algal blooms. In this study, critical (optimal) conditions and mechanisms for AlCl

Purpose: Compressed Sensing Magnetic Resonance Imaging (CS-MRI) is a promising technique to accelerate dynamic cardiac MR imaging (DCMRI). For DCMRI, the CS-MRI usually exploits image signal sparsity and low-rank property to reconstruct dynamic images from the undersampled k-space data. In this paper, a novel CS algorithm is investigated to improve dynamic cardiac MR image reconstruction quality under the condition of minimizing the k-space recording.Methods: The sparse representation of 3D cardiac magnetic resonance data is implemented by synergistically integrating 3D TGV algorithm and high order singular value decomposition (HOSVD) based Tensor Decomposition, termed as k-t TGV-TD method. In the proposed method, the low rank structure of the 3D dynamic cardiac MR data is performed by the HOSVD method, and the localized image sparsity is achieved by the 3D TGV method. Moreover, the Fast Composite Splitting Algorithm (FCSA) method, combining the variable splitting with operator splitting techniques, is employed to solve the low-rank and sparse problem. Two different cardiac MR datasets (cardiac cine and cardiac perfusion MR data) are used to evaluate the performance of the proposed method.Results: Compared with the state-of-art methods, such as the k-t SLR method, 3D TGV method and HOSVD based tensor decomposition method, the proposed method can offer improved reconstruction accuracy in terms of higher signal-to-error ratio (SER).Conclusions: This work proved that the k-t TGV-TD method was an effective sparse representation way for DC-MRI, which was capable of significantly improving the reconstruction accuracy with different reduction factor.

In this study, disinfection by-products (DBP) formation from dissolved organic matter (DOM) and its fractions, including both hydrophilic and hydrophobic components, were investigated at a typical karst surface water. The subsequent DBP formation potential was evaluated by deducing chemical characteristics of DOM fractions and representative algal organic matter (Chlorella sp. AOM) under the influence of divalent ions (Ca

Osseointegration is the direct and intimate contact between mineralized tissue and titanium implant at the bone-implant interface. Early establishment and stable maintenance of osseointegration is the key to long-term implant success. However, in patients with compromised conditions such as osteoporosis and patients beginning early load-bearing activities such as walking, lower osseointegration around titanium implants is often observed, which might result in implant early failure. Gene-activated implants show an exciting prospect of combining gene delivery and biomedical implants to solve the problems of poor osseointegration formation, overcoming the shortcomings of protein therapy, including rapid degradation and overdose adverse effects. The conception of gene-activated titanium implants is based on "gene-activated matrix" (GAM), which means scaffolds using non-viral vectors for in situ gene delivery to achieve a long-term and efficient transfection of target cells. Current preclinical studies in animal models have shown that plasmid DNA (pDNA), microRNA (miRNA), and small interference RNA (siRNA) functionalized titanium implants can enhance osseointegration with safety and efficiency, leading to the expectation of applying this technique in dental and orthopedic clinical scenarios. This review aims to comprehensively summarize fabrication strategies, current applications, and futural outlooks of gene-activated implants, emphasizing nucleic acid targets, non-viral vectors, implant surface modification techniques, nucleic acid/vector complexes loading strategies.

Carbon nanomaterials, such as carbon nanoonions (CNOs), possess promising applications in various fields. There are urgent demands to synthesize carbon nanomaterials from a green and renewable carbon source. In this study, solid CNOs with relatively uniform size distribution (with diameters of about 30-50 nm), abundant structure defects and oxygen-containing surface functional groups (such as -OH and -COOH) are developed from co-pyrolysis of lignin (LG) and polyethylene (PE) in the presence of Ni-based catalysts. The type of catalyst, the concentration of catalyst and catalytic co-pyrolysis temperature play important roles in the morphologies and properties of CNOs as confirmed by TEM and SEM. Furthermore, the produced CNOs can act as a low-cost and highly-efficient adsorbent to remove Cu(ii) from aqueous solution according to a homogeneous monolayer, chemical action-dominated, endothermic and spontaneous process. The theoretical maximum adsorption capacity of CNOs calculated from the Langmuir model is 100.00 mg g