Search

Social networks allow users to actively upload images and descriptive tags, which has led to an explosive growth in the number of social images. Multi-view hashing is an efficient technique for supporting large-scale social image retrieval because of its desirable capabilities of encoding multi-view features into compact binary hash codes with extremely low storage costs and fast retrieval speeds. However, existing methods require multi-view features to be fully paired at both the offline model training and online query stages. This requirement cannot be easily satisfied for social image retrieval, where social images that lack descriptive tags are common in social networks. In this paper, we propose an Unsupervised Adaptive Partial Multi-view Hashing (UAPMH) method to handle the partial-view hashing problem for efficient social image retrieval. Specifically, the shared and view-specific latent representations of fully paired and partial-view images, respectively, are learned separately by an adaptive partial multi-view matrix factorization module within the identical semantic space. In particular, instead of adopting simple fixed view combination weights, we develop a parameter-free weight learning scheme to adaptively learn the weights to capture the view variations and the discriminative capabilities of different views. With such a design, our model can sufficiently exploit the available partial-view samples with separate hash code learning and effectively preserve the latent relations of images and tags in hash codes with semantic space sharing. Moreover, to avoid relaxing errors and improve the learning efficiency, binary hash codes are directly learned in a fast mode with simple and efficient operations. Finally, we extend UAPMH to the supervised learning paradigm as Supervised Adaptive Partial Multi-view Hashing (SAPMH) with the supervision of pair-wise semantic labels to further enhance the discriminative capability of hash codes. The experiments demonstrate the state-of-the-art performance of the proposed approaches on public social image retrieval datasets. Our source codes and testing datasets can be obtained at https://github.com/ChaoqunZheng/APMH .

Learning a second language (L2) after puberty is well-known to be subject to a great\ud deal of individual variation. Even if two individuals spend the same amount of time\ud practicing a target language, their final outcomes may differ greatly. This could be in part due\ud to the fact that certain individuals are perceptually and cognitively adept at making the most\ud of every practice opportunity (i.e., they possess higher L2 learning aptitude), resulting in\ud more advanced L2 proficiency (Doughty, 2019). Whereas scholars have extensively\ud examined which perceptual-cognitive abilities relate to successful L2 learning, most of the\ud existing literature has been exclusively concerned with lexicogrammar aspects of language\ud learning (Li, 2016). In light of the ongoing discussion regarding the similarities between\ud speech, music and language learning (Tierney, Krizman, & Kraus, 2015), we highlight two\ud overlapping abilities, music aptitude and auditory processing, as a framework of aptitude\ud relevant to successful L2 pronunciation learning. Music aptitude is defined as a set of\ud composite, domain-specific abilities to remember and reproduce music phrases that are no\ud longer physically present, generally measured through standardized tests, and found to relate\ud to L2 pronunciation development to some degree (Slevc & Miyake, 2006). More specifically,\ud scholars have been interested in domain-general sensitivity to more fine-grained properties of\ud acoustic signals (formants, pitch, duration, amplitude), which we refer to here collectively as\ud auditory processing. The ability has been linked to first language (L1) acquisition (e.g.,\ud Goswami, 2015), and to L2 acquisition (e.g., Kachlicka, Saito, & Tierney, 2019). To test\ud whether music aptitude and auditory processing abilities explain variation in adult L2\ud pronunciation learning, we investigated the complex relationship between music aptitude,\ud auditory processing, and biographical profiles of 48 Chinese learners of English in the UK.

The purpose of this study was to explore the significance of the neuregulin-1/ErbB signaling pathway and its effect on Sox10 expression in the course of the differentiation of mouse bone marrow mesenchymal stem cells into Schwann-like cellsThe experiment was conducted with three groups-control, TAK 165, and HRG-off. In the control group, we used the classical induction method of adding β-ME, RA, FSK, b-FGF, PDGF, and neuregulin (HRG); the cells were collected on the 7Our results showed that the proliferation of Schwann cells was reduced and apoptosis was increased in the TAK 165 group and the HRG-off group. Sox10 was stably expressed and NRG-1, ErbB2, and ErbB3 increased in the control group. However, the expression of Sox10 in the TAK 165 group was obviously decreased at the end of induced differentiation; meanwhile, the degree of stem cell differentiation also decreased.the neuregulin-1/ErbB signaling pathway plays an important role in the differentiation of bone marrow mesenchymal stem cells into Schwann-like cells and can promote the maintenance of Sox10 。.

Background: Low plasma potassium (p-K) is associated with increased risk of malignant arrhythmia and observational studies indicate protective effects of p-K in the upper reference level. However, randomized clinical studies are needed to document whether actively increasing p-K to high-normal levels is possible and safe and improves cardiovascular outcomes. Objective: To investigate if increased p-K reduces the risk of malignant arrhythmia and all-cause death in high-risk patients with a cardiovascular disease treated with an implantable cardioverter defibrillator (ICD) for primary or secondary preventive causes. Secondly, to investigate whether high-normal p-K levels can be safely reached and maintained using already available medications and potassium-rich dietary guidance. Methods: This is a prospective, randomized, and open-labelled study enrolling patient at high-risk of malignant arrhythmias. According to sample size calculations, 1,000 patients will be randomized 1:1 to either an investigational regiment that aims to increase and maintain p-K at high-normal levels (4.5-5.0 mmol/L) or to usual standard of care and followed for an expected four years. The trial will run until a total of 291 events have occurred providing an α = 0.05 and 1-β = 0.80. The composite primary endpoint includes ventricular tachycardia >125 bpm lasting >30 seconds, any appropriate ICD-therapy, and all-cause mortality. At present, 739 patients have been randomized. Conclusions: We present the rationale for the design of the POTCAST trial. The inclusion was initiated 2019 and is expected to be finished 2022. The study will show if easily available treatments to increase p-K may be a new treatment modality to protect against malignant arrythmias.

Given that Delay Constrained Relay Node Placement (DCRNP) problem is getting more and more attention in time-critical domains, this paper proposes an Improved Greedy-based Selection Algorithm (IGSA) to tackle such problem under delay and budget constraints. IGSA first employs communication cost matrix and Dijkstra’s algorithm to detect whether the problem is solvable, then selects the amount and locations of deployed relay nodes (RNs) layer by layer by continuously calling the greedy coverage algorithm until all sensors have a feasible path to the sink under a preset delay constraint, and finally removes redundant RNs to reach the approximate optimal solution. Extensive simulations show that our approach can save 15%–30% of deployed RNs than the Two-phase Set-Covering-based Algorithm (TSCA), which is the state-of-the-art algorithm for the DCRNP problem, only with a slight sacrifice of solving time. Meanwhile, Ad Hoc On-Demand Distance Vector (AODV) routing protocol is applied to further examine the performance of the deployment results solved by TSCA and IGSA, respectively. Results show that every RN deployed by IGSA is involved in data forwarding, while IGSA deploys leisure RNs which do not participate in the data forwarding. Besides, compared with TSCA, the network formed by IGSA has lower network delay in the data transmission process, and the energy consumption of the central node and throughput of the two topologies are almost at the same level.

Adult zebrafish (Danio rerio) were electrolyzed at different current densities to explore the effects of electrolysis on their reproductive system, especially on embryo production, and to uncover the molecular mechanism of changes in sex hormone and vitellogenin (VTG) levels. The results showed that embryo reproduction of zebrafish was reduced at a current density of 0.64 A/m

Recently, electrolysis technology has been widely applied in nitrogen and phosphorus removal in river water due to its high efficiency, but its effects on aquatic animals, especially on their neurodevelopmental system, are still unclear. In this study, zebrafish (Danio rerio) embryos were used as model organisms and were put into an electrolytic reaction device with a Ti/IrO

The application of electric actuator in aircraft is increasing due to the concept of more electric aircraft. The electrical trimmable horizontal stabilizer actuator (ETHSA) is also a type of electric actuator, which is a significant part of the aircraft flight control system. There has been a lot of research on the modeling of ETHSA, but few of them focus on the fault detection and isolation(FDI). In this article, the bond graph model of each component is derived and the full model of ETHSA is obtained firstly. Then establish diagnostic model, analyze the redundancy relations and building the fault signature matrix. Simulation results verify the FDI of ETHSA is feasible based on 20-sim simulation environment.

Unsupervised multi-modal hashing has recently attracted broad attention in research area of large-scale multimedia retrieval for its low storage cost, high retrieval speed, and independence on semantic labels. However, the model learning process of existing methods still suffer from the problem of low efficiency: 1) Many existing methods measure the contributions of different modalities using fixed modality weights. In order to avoid over-fitting, they need an inefficient hyper-parameter adjustment process. 2) Most existing methods adopt inefficient optimization strategies to learn hash codes. In this letter, we propose an unsupervised Efficient Parameter-free Adaptive Multi-modal Hashing (EPAMH) model to adaptively capture the modality variations and preserve the discriminative semantics of multi-modal features into the binary hash codes. Moreover, we directly learn the binary codes with simple and efficient operations, which prevents the relaxing quantization errors and improves the model learning efficiency. Experiments prove the superior performance of EPAMH on three public multimedia retrieval datasets. Our source codes and testing datasets can be obtained at https://github.com/ChaoqunZheng/EPAMH .

In order to testify the effect of electrolysis and microbial remediation technology in polluted river sediment. Here, we explored the possibility of electrochemically removing ammoniacal nitrogen-nitrogen (NH3-N), nitrate-nitrogen (NO3−-N) and phosphate ions-phosphorous (PO43−-P) by using a titanium (Ti) mesh cathode, a Ti/Ti dioxide (TiO2)/Ruthenium (IV) oxide (RuO2) (RuO2-IrO2/Ti) mesh, and a magnesium-aluminum (Mg–Al) alloy anode placed within the sediment and overlying water. Results showed that approximately 151.82 ± 21.69 mg TN was removed which was five times more effective than the non-electrolytic controls (30.21 ± 13.73 mg), NH3-N concentration in the sediment was substantially reduced (up to 2.9 times) compared to the non-electrolytic controls. Its efficiency lies in the electrolysis process, which may directly remove NH3-N through electrochemical oxidation and simultaneously produce oxygen which helps nitrifying bacteria to convert NH3-N into NO3−-N by the role of anode; and electrolysis may directly remove NO3−-N in the overlying water through electrochemical reduction while simultaneously producing hydrogen electron donor for hydrogen autotrophic microorganism as Hydrogenophohaga, to be the dominant species in sediment to enhance the removal of NO3−-N by the role of cathode. Electrolysis also reduced the PO43−-P through electro-coagulation since Mg2+ ions could also produce since sacrificial Mg–Al alloy anode was used and electro-deposition on Ti mesh cathode both to increase PO43−-P removal in overlying water and sediment. This study verifies the benefits of electrolysis-driven bioremediation as a sustainable technology for the bioremediation of N and P polluted river sediments.

Supplemental material, sj-docx-2-slr-10.1177_0267658320978493 for Successful second language pronunciation learning is linked to domain-general auditory processing rather than music aptitude by Chaoqun Zheng, Kazuya Saito and Adam Tierney in Second Language Research

In this paper, a simple method is proposed for detecting salient regions by utilizing muti-scale features at a superpixel-based level. To guarantee the accuracy of the detected region, we make use of superpixel to form a new measurement, instead of using the level of pixel. Majority of saliency models consider visual features such as orientation, color, intensity and the importance of visual features has not yet been fully explored. Here, we take texture feature into consideration in particular, using the gray level co-occurrence matrix(GLCM) as a texture descriptor. Specifically, a new local smoothing filter based on GLCM is proposed to smooth image and remove some simple texture before getting GLCM. Experimental results demonstrate that our approach significantly outperforms some existing popular methods with full resolution and well-defined boundary.

Nurturing critical thinking (CT) has been acknowledged as a core objective of tertiary education, and drawn attention from academia of teaching English as a Foreign Language (EFL) in China. The thrust of the present study was to determine the association between CT and EFL argumentative writing among Chinese undergraduates. To this end, 110 English majors across three grades at two universities were conveniently selected and given the critical thinking skills (CTS) test and EFL argumentative writing test. The results of this study indicated that undergraduate English majors in China did not possess strong CTS. Though their CTS was not found to be significantly correlated with EFL argumentative writing performance, textual analysis of typical essays showed that strong-CTS learners outperformed weak-CTS ones in relevance, clarity, logicality, profundity and flexibility of argumentative writing. The obtained results suggest a need to integrate CT into EFL writing instruction.