A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2021; you can also visit the original URL.
The file type is
Instance segmentation on point clouds is a fundamental task in 3D scene perception. In this work, we propose a concise clustering-based framework named HAIS, which makes full use of spatial relation of points and point sets. Considering clustering-based methods may result in over-segmentation or under-segmentation, we introduce the hierarchical aggregation to progressively generate instance proposals, i.e., point aggregation for preliminarily clustering points to sets and set aggregation forarXiv:2108.02350v1 fatcat:d6hvyhtfyvc4jkhcvo7lzeyryi
more »... erating complete instances from sets. Once the complete 3D instances are obtained, a sub-network of intra-instance prediction is adopted for noisy points filtering and mask quality scoring. HAIS is fast (only 410ms per frame) and does not require non-maximum suppression. It ranks 1st on the ScanNet v2 benchmark, achieving the highest 69.9% AP50 and surpassing previous state-of-the-art (SOTA) methods by a large margin. Besides, the SOTA results on the S3DIS dataset validate the good generalization ability. Code will be available at https://github.com/hustvl/HAIS.
Human pose estimation from image and video is a vital task in many multimedia applications. Previous methods achieve great performance but rarely take efficiency into consideration, which makes it difficult to implement the networks on resource-constrained devices. Nowadays real-time multimedia applications call for more efficient models for better interactions. Moreover, most deep neural networks for pose estimation directly reuse the networks designed for image classification as the backbone,arXiv:2012.07086v1 fatcat:bc5o6lzfvrgevgghacsfxgjaay
more »... which are not yet optimized for the pose estimation task. In this paper, we propose an efficient framework targeted at human pose estimation including two parts, the efficient backbone and the efficient head. By implementing the differentiable neural architecture search method, we customize the backbone network design for pose estimation and reduce the computation cost with negligible accuracy degradation. For the efficient head, we slim the transposed convolutions and propose a spatial information correction module to promote the performance of the final prediction. In experiments, we evaluate our networks on the MPII and COCO datasets. Our smallest model has only 0.65 GFLOPs with 88.1% PCKh@0.5 on MPII and our large model has only 2 GFLOPs while its accuracy is competitive with the state-of-the-art large model, i.e., HRNet with 9.5 GFLOPs.
TELKOMNIKA Indonesian Journal of Electrical Engineering
Since there has been a strong demand from industry to have an efficient way of managing color image quality presented in different media, by specifically investigating partly changed complex images, this article proposed a revision to existing CIE color difference model which cannot give a proper color difference assessment on partly changed complex images. The key method applied is to find out weight coefficients of color attributes such as lightness, hue and chroma in color difference prediction.doi:10.11591/telkomnika.v12i11.6035 fatcat:n6rboiqj75h5vark5t2ahob77u
Radiation-induced lung injury (RILI) is a common complication of thoracic radiotherapy, but efficacious therapy for RILI is lacking. This study ascertained whether glycyrrhetinic acid (GA; a functional hydrolyzed product of glycyrrhizic acid, which is extracted from herb licorice) can protect against RILI and investigated its relationship to the transforming growth factor (TGF)-β1/Smads signaling pathway. C57BL/6 mice were divided into four groups: a control group, a GA group and twodoi:10.1093/jrr/rrw091 pmid:27672101 pmcid:PMC5321194 fatcat:ybaf6zxlr5a5rod7a2nb23niyy
more »... (IR) groups. IR groups were exposed to a single fraction of Xrays (12 Gy) to the thorax and administered normal saline (IR + NS group) or GA (IR + GA group). Two days and 17 days after irradiation, histologic analyses were performed to assess the degree of lung injury, and the expression of TGF-β1, Smad2, Smad3 and Smad7 was recorded. GA administration mitigated the histologic changes of lung injury 2 days and 17 days after irradiation. Protein and mRNA expression of TGF-β1, Smad2 and Smad3, and the mRNA level of Smad7, in lung tissue were significantly elevated after irradiation. GA decreased expression of TGF-β1, Smad2 and Smad3 in lung tissue, but did not increase Smad7 expression. GA can protect against earlystage RILI. This protective effect may be associated with inhibition of the TGF-β1/Smads signaling pathway.
With the popularity of music, the music equipment market has ushered in a new round of explosion. But at the same time, the market environment is changing rapidly, and music equipment companies are greatly affected by market changes. This type of enterprise has a wide variety of products, fast update, short delivery time and urgent time, which is a severe challenge to its production and operation. In this environment, in order to quickly respond to changes in the market environment, and todoi:10.1155/2022/6580742 doaj:22baf27c588644cda0c5742e311bb681 fatcat:bvfqqayhxnagvjwt437pv7eqae
more »... late plans for corporate procurement, production, and sales in an orderly manner, music equipment companies must make accurate predictions of market demand. At present, the extended research based on LSTM is a relatively mainstream deep neural network in the research methods of time series problems. This article is based on LSTM model to conduct an in-depth study on the prediction of music equipment demand. In order to solve the problems of overfitting, disappearance of gradient, model collapse and other problems in previous experimental studies, this paper proposes an improved LSTM prediction model. In terms of model structure selection, Dropout mechanism is used, and L2 regular term is introduced. In the selection of the activation function, MReLU function is proposed, which can improve the prediction effect of the model and enhance the applicability of the model. To measure the prediction effect of the improved LSTM model established in this paper, this paper selects RMSE and MAE as evaluation indicators, and compares experiments with other mainstream prediction models. The research results show that the improved LSTM network prediction model is superior to other models in the prediction of music equipment demand, which verifies the effectiveness.
., Jiemin Zhang and Jephias Gwamuri performed the experiments; all authors analyzed the data, wrote and edited the paper. Conflicts of Interest: The authors declare no conflict of interest. ...doi:10.3390/en10081194 fatcat:lin4n56qwbefrnjgcwznry3qhu
In this study, ovalbumin (OVA) was succinylated with the addition of different levels of succinic anhydride, and the structural and functional properties of succinylated OVA (SOVA) were investigated. SDS−PAGE and FTIR spectrum confirmed the covalent attachment of the succinyl group to OVA. Thermal stability and the absolute value of zeta potential (pH 6.0) of SOVA were enhanced by 14.90% and 76.77% higher than that of the native OVA (NOVA), respectively. Circular dichroism (CD) spectradoi:10.3390/foods11182724 pmid:36140852 pmcid:PMC9497935 fatcat:4qe3v3hm4bawpk5ep6vqaqb3gu
more »... ted that the succinylation decreased the α−helix and increased β−sheet content to 21.31% and 43.28%, respectively. The content of free sulfhydryl groups increased and intrinsic fluorescence spectra suggested the SOVA became more unfolded and flexible as the degree of succinylation enhanced. Furthermore, succinylation effectively enhanced the solubility and decreased the interface tension (oil−water and air−water interface) of OVA. Compared to NOVA, the emulsifying activity and stability of SOVA were increased by 1.6 times and 1.2 times, respectively, and foaming capacity and stability were enhanced by 2.7 times and 1.5 times, respectively.
AbstractHuman pose estimation from image and video is a key task in many multimedia applications. Previous methods achieve great performance but rarely take efficiency into consideration, which makes it difficult to implement the networks on lightweight devices. Nowadays, real-time multimedia applications call for more efficient models for better interaction. Moreover, most deep neural networks for pose estimation directly reuse networks designed for image classification as the backbone, whichdoi:10.1007/s41095-021-0214-z fatcat:3vgbohaew5bavgbpjcsq3tcxji
more »... re not optimized for the pose estimation task. In this paper, we propose an efficient framework for human pose estimation with two parts, an efficient backbone and an efficient head. By implementing a differentiable neural architecture search method, we customize the backbone network design for pose estimation, and reduce computational cost with negligible accuracy degradation. For the efficient head, we slim the transposed convolutions and propose a spatial information correction module to promote the performance of the final prediction. In experiments, we evaluate our networks on the MPII and COCO datasets. Our smallest model requires only 0.65 GFLOPs with 88.1% PCKh@0.5 on MPII and our large model needs only 2 GFLOPs while its accuracy is competitive with the state-of-the-art large model, HRNet, which takes 9.5 GFLOPs.
Due to the development of XML and other data models such as OWL and RDF, sharing data is an increasingly common task since these data models allow simple syntactic translation of data between applications. However, in order for data to be shared semantically, there must be a way to ensure that concepts are the same. One approach is to employ commonly used schemas -called standard schemas -which help guarantee that syntactically identical objects have semantically similar meanings. As a resultdoi:10.1016/j.is.2010.08.005 fatcat:5dgkuwhdirgtzht7i4snzim4se
more »... the spread of data sharing, there has been widespread adoption of standard schemas in a broad range of disciplines and for a wide variety of applications within a very short period of time. However, standard schemas are still in their infancy and have not yet matured or been thoroughly evaluated. It is imperative that the data management research community takes a closer look at how well these standard schemas have fared in real-world applications to identify not only their advantages, but also the operational challenges that real users face. In this paper, we both examine the usability of standard schemas in a comparison that spans multiple disciplines, and describe our first step at resolving some of these issues in our Semantic Modeling System. We evaluate our Semantic Modeling System through a careful case study of the use of standard schemas in Architecture, Engineering, and Construction, which we conducted with domain experts. We discuss how our Semantic Modeling System can help the broader problem and also discuss a number of challenges that still remain.
Our previous studies demonstrated that tentacle extract (TE) from the jellyfish, Cyanea capillata, could cause a dose-dependent increase of systolic blood pressure, which seemed to be the result of direct constriction of vascular smooth muscle (VSM). The aim of this study is to investigate whether TE could induce vasoconstriction in vitro and to explore its potential mechanism. Using isolated aorta rings, a direct contractile response of TE was verified, which showed that TE could inducedoi:10.3390/md11093335 pmid:23999662 pmcid:PMC3806464 fatcat:54wum4sslnecxaqykwklgfpayy
more »... ration-dependent contractile responses in both endothelium-intact and -denuded aortas. Interestingly, the amplitude of contraction in the endothelium-denuded aorta was much stronger than that in the endothelium-intact one, implying that TE might also bring a weak functional relaxation in addition to vasoconstriction. Further drug intervention experiments indicated that the functional vasodilation might be mediated by nitric oxide, and that TE-induced vasoconstriction could be attributed to calcium influx via voltage-operated calcium channels (VOCCs) from the extracellular space, as well as sarcoplasmic reticulum (SR) Ca 2+ release via the inositol 1,4,5-trisphosphate receptor (IP 3 R), leading to an increase in [Ca 2+ ] c , instead of activation of the PLC/DAG/PKC pathway or the sympathetic nerve system.
With the development of large-scale CMOS-integrated circuit manufacturing technology, microprocessor chips are more vulnerable to soft errors and radiation interference, resulting in reduced reliability. Core reliability is an important element of the microprocessor's ability to resist soft errors. This paper proposes DuckCore, a fault-tolerant processor core architecture based on the free and open instruction set architecture (ISA) RISC-V. This architecture uses improved SECDED (single errordoi:10.3390/electronics11010122 fatcat:sgsfomh3ojfxhocl67gbrnxk2i
more »... rrection, double error detection) code between pipelines, detects processor operating errors in real-time through the Supervision unit, and takes instruction rollbacks for different error types, which not only saves resources but also improves the reliability of the processor core. In the implementation process, all error injection tests are passed to verify the completeness of the function. In order to better verify the performance of the processor under different error intensity injections, the software is used to inject errors, the running program is run on the FPGA (Field Programmable Gate Array), and the impact of the actual radiation environment on the architecture is evaluated through the results. The architecture is applied to three–five-stage open-source processor cores and the results show that this method consumes fewer resources and its discrete design makes it more portable.
Ferroportin is the only cellular iron exporter in human and essential for iron homoeostasis. Mutations in ferroportin cause ferroportin diseases characterized by a paradoxical combination of anemia and abnormal accumulation of iron in cells. Ferroportin is also the target of hepcidin, which is a hormone that downregulates ferroportin activity. However, due to a lack of three-dimensional structures, the mechanism of iron transport in ferroportin and its regulation by hepcidin remains unclear.doi:10.1101/2020.03.04.975748 fatcat:7yj54zh2ebg55liufiwlysh7xy
more »... e we present the structure of a ferroportin from the primate Tarsius syrichta (TsFpn) at 3.0 angstrom resolution solved by cryo-electron microscopy. TsFpn has a structural fold common to major facilitator superfamily of transporters and the current structure is in an outward-open conformation. The structure identifies two potential ion binding sites and each site is coordinated by two residues. Functional studies demonstrate that TsFpn is a H+/Fe2+ antiporter and that transport of one Fe2+ is coupled to the transport of two H+ in the opposite direction so that the transport cycle is electroneutral. Mutation to one of the sites mainly affect H+ transport while mutation to the other site affects both Fe2+ and H+ transport. The structure also provides mechanistic interpretation for mutations that cause ferroportin diseases.
Underwater ghost imaging based on deep learning can effectively reduce the influence of forward scattering and back scattering of water. With the help of data-driven methods, high-quality results can be reconstructed. However, the training of the underwater ghost imaging requires enormous paired underwater datasets, which are difficult to obtain directly. Although the Cycle-GAN method solves the problem to some extent, the blurring degree of the fuzzy class of the paired underwater datasetsdoi:10.3390/s22166161 pmid:36015921 pmcid:PMC9412451 fatcat:ukjvi56vj5h4dl4rk2ypybh7ne
more »... rated by Cycle-GAN is relatively unitary. To solve this problem, a few-shot underwater image generative network method is proposed. Utilizing the proposed few-shot learning image generative method, the generated paired underwater datasets are better than those obtained by the Cycle-GAN method, especially under the condition of few real underwater datasets. In addition, to reconstruct high-quality results, an underwater deblurring ghost imaging method is proposed. The reconstruction method consists of two parts: reconstruction and deblurring. The experimental and simulation results show that the proposed reconstruction method has better performance in deblurring at a low sampling rate, compared with existing underwater ghost imaging methods based on deep learning. The proposed reconstruction method can effectively increase the clarity degree of the underwater reconstruction target at a low sampling rate and promotes the further applications of underwater ghost imaging.
This paper deals with the problem of constant false alarm rate (CFAR) target detection in high-resolution ground synthetic aperture radar (SAR) images based on KK distribution. For the parameter estimation of KK distribution, the semi-experiential algorithm is analyzed firstly. Then a new estimation algorithm based on the particle swarm optimization (PSO) is proposed, which takes the discrepancies between the histogram of the clutter data and probability density function (PDF) of KKdoi:10.2528/pier13031602 fatcat:ggch7b3ueveqlpobvcevxuehae
more »... at some selected points as the cost function to search for the optimal parameter values using PSO algorithm. The performance of the two algorithms is compared using Monte-Carlo simulation using the simulated data sets generated under different conditions; and the estimation results validate the better performance of the new algorithm. Then the KK distribution, which is proposed for spiky sea clutter originally, is applied to model the real ground SAR clutter data. The goodness-of-fit test clearly show that the KK distribution is able to model the ground SAR clutter much better than some common used model, such as standard Kdistribution and Gamma, etc. On this basis, a global CFAR target detection algorithm is presented. The detection threshold is calculated numerically through the cumulative density function (CDF) of KK distribution. Comparing the amplitude of every SAR image pixel with this threshold, the potential targets in ground SAR images can be located effectively. Then target clustering is implemented to eliminate the false alarm and obtain more accurate target regions. The detection results of the proposed algorithm in a typical ground SAR image show that it has better performance than the detector based on G 0 distribution.
Zhang designed the experiments and revised the paper. All authors discussed the results and reviewed the manuscript. ... Zhang conducted the experiments and wrote the paper; J.M. Li and X. Yang analyzed the results; J.M. Li and R.L. Zhu performed the corrosion and noise tests; X. Yang and Y. ...doi:10.3390/met4040597 fatcat:qxtkj6f2mncqdjrlmmbfjd3sfi
« Previous Showing results 1 — 15 out of 404 results