A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf
.
Filters
Pyramid Mask Text Detector
[article]
2019
arXiv
pre-print
Scene text detection, an essential step of scene text recognition system, is to locate text instances in natural scene images automatically. Some recent attempts benefiting from Mask R-CNN formulate scene text detection task as an instance segmentation problem and achieve remarkable performance. In this paper, we present a new Mask R-CNN based framework named Pyramid Mask Text Detector (PMTD) to handle the scene text detection. Instead of binary text mask generated by the existing Mask R-CNN
arXiv:1903.11800v1
fatcat:zleabbqd3fcznhsok5dzx5tosi
more »
... ed methods, our PMTD performs pixel-level regression under the guidance of location-aware supervision, yielding a more informative soft text mask for each text instance. As for the generation of text boxes, PMTD reinterprets the obtained 2D soft mask into 3D space and introduces a novel plane clustering algorithm to derive the optimal text box on the basis of 3D shape. Experiments on standard datasets demonstrate that the proposed PMTD brings consistent and noticeable gain and clearly outperforms state-of-the-art methods. Specifically, it achieves an F-measure of 80.13% on ICDAR 2017 MLT dataset.
Difficulty-Aware Machine Translation Evaluation
[article]
2021
arXiv
pre-print
There are many variants of our proposed method: 1) designing more elaborate difficulty function (Liu et al., 2020; Zhan et al., 2021a) ; 2) applying a smoothing function to the difficulty distribution ...
arXiv:2107.14402v1
fatcat:bhpdvyn4ajfspailterr5rcpzm
Variance-Aware Machine Translation Test Sets
[article]
2021
arXiv
pre-print
., 2021a) or representation enhancement techniques (Nguyen and Chiang, 2018; Liu et al., 2019) , resulting in the differences of lexical choice performed on the test set. ...
arXiv:2111.04079v1
fatcat:gf43v57z7vdnhon3h7qk2pf4ly
Norm-Based Curriculum Learning for Neural Machine Translation
[article]
2020
arXiv
pre-print
(vector) can be divided into two parts: the norm and the direction: w = ||w|| norm · w ||w|| direction (5) In practice, the word embedding, represented by w, is the key component of a neural model (Liu ...
arXiv:2006.02014v1
fatcat:e54jamzdpbe7tk7lwyga5irvyy
Simultaneous Ultrasound and Heat Enhance Functional Properties of Glycosylated Lactoferrin
2020
Molecules
Protein-polysaccharide covalent complexes exhibit better physicochemical and functional properties than single protein or polysaccharide. To promote the formation of the covalent complex from lactoferrin (LF) and beet pectin (BP), we enhanced the Maillard reaction between LF and BP by using an ultrasound-assisted treatment and studied the structure and functional properties of the resulting product. The reaction conditions were optimized by an orthogonal experimental design, and the highest
doi:10.3390/molecules25235774
pmid:33297553
pmcid:PMC7730217
fatcat:cxfekvasdnhrhd7lkuobqw7v5q
more »
... ting degree of 55.36% was obtained by ultrasonic treatment at 300 W for 20 min and at LF concentration of 20 g/L and BP concentration of 9 g/L. The formation of LF-BP conjugates was confirmed by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and Fourier transform infrared (FTIR) spectroscopy. Ultrasound-assisted treatment can increase the surface hydrophobicity, browning index, 1,1-diphenyl-2-picryl-hydrazyl (DPPH) and 2,2'-azinobis-(3-ethylbenzthiazoline-6-sulphonate) (ABTS) free radicals scavenging activity of LF due to the changes in the spatial configuration and formation of Maillard reaction products. The thermal stability, antioxidant activity and emulsifying property of LF were significantly improved after combining with BP. These findings reveal the potential application of modified proteins by ultrasonic and heat treatment.
Physicochemical Properties of Lutein-Loaded Microcapsules and Their Uptake via Caco-2 Monolayers
2018
Molecules
Lutein is one of the most important carotenoids that can be utilized in foods as a natural pigment and nutraceutical ingredient to improve eye health. However, its utilization is limited due to its poor solubility. Chemically, the highly unsaturated structure of lutein makes it extremely susceptible to light, oxygen, heat, and pro-oxidants and therefore easily oxidized, decomposed or dissociated. In this study, we aimed to imbed natural lutein to improve its storage stability and enhance its
doi:10.3390/molecules23071805
pmid:30037053
fatcat:26l6y6qxpnhrnjq3g2galduiau
more »
... er dispersibility. As two commonly studied water-soluble and water-insoluble food-grade surfactants, lecithin and sodium caseinate (NaCas) were chosen as the wall materials, and lutein-loaded lecithin microcapsules and NaCas microcapsules were prepared, the results revealed the lutein-loaded NaCas microcapsules not only exhibited better solubility and stability than those of lutein-loaded lecithin microcapsules, but also were more stable when stored at 4 °C, 25 °C, 37 °C. Moreover, the lutein-loaded NaCas microcapsules were more easily absorbed by the intestinal Caco-2 cells than natural lutein. Considering the dispersibility, stability and cell absorption effect, the NaCas-based microparticle is a potential carrier for lutein.
Methionine restriction alleviates age-associated cognitive decline via fibroblast growth factor 21
2021
Redox Biology
Zhigang Liu is also funded by the Tang Cornell-China Scholars Program from Cornell University in the U.S. and the Alexander von Humboldt-Stiftung in Germany. ...
doi:10.1016/j.redox.2021.101940
pmid:33765615
pmcid:PMC8022247
fatcat:d2tywap3ibacbm4yfzrjh3wmam
Fermentation of tomato juice improves in vitro bioaccessibility of lycopene
2020
Journal of Functional Foods
Yuyan Lu and Fuguo Liu analyzed the data and wrote the manuscript. David Julian McClements, Xuebo Liu and Fuguo Liu reviewed and revised the manuscript. ...
Yuyan Lu and Fuguo Liu conceived and designed the experiments. Yuyan Lu, Kaiyu Mu and Xiuping Liang performed the experiments. Xuebo Liu contributed helpful discussion during the experiment. ...
doi:10.1016/j.jff.2020.104020
fatcat:nbs64oq3unagvj7ki7fxsezqmq
PV-EV Integrated Home Energy Management Considering Residential Occupant Behaviors
2021
Sustainability
Rooftop photovoltaics (PV) and electrical vehicles (EV) have become more economically viable to residential customers. Most existing home energy management systems (HEMS) only focus on the residential occupants' thermal comfort in terms of indoor temperature and humidity while neglecting their other behaviors or concerns. This paper aims to integrate residential PV and EVs into the HEMS in an occupant-centric manner while taking into account the occupants' thermal comfort, clothing behaviors,
doi:10.3390/su132413826
fatcat:zo2n3ofkpvh4lmuvghxvqtsyju
more »
... d concerns on the state-of-charge (SOC) of EVs. A stochastic adaptive dynamic programming (ADP) model was proposed to optimally determine the setpoints of heating, ventilation, air conditioning (HVAC), occupant's clothing decisions, and the EV's charge/discharge schedule while considering uncertainties in the outside temperature, PV generation, and EV's arrival SOC. The nonlinear and nonconvex thermal comfort model, EV SOC concern model, and clothing behavior model were holistically embedded in the ADP-HEMS model. A model predictive control framework was further proposed to simulate a residential house under the time of use tariff, such that it continually updates with optimal appliance schedules decisions passed to the house model. Cosimulations were carried out to compare the proposed HEMS with a baseline model that represents the current operational practice. The result shows that the proposed HEMS can reduce the energy cost by 68.5% while retaining the most comfortable thermal level and negligible EV SOC concerns considering the occupant's behaviors.
Identification of Regulatory circRNAs Involved in the Pathogenesis of Acute Myocardial Infarction
2021
Frontiers in Genetics
BackgroundAcute myocardial infarction (AMI) has high morbidity and mortality worldwide. However, the pathogenesis of AMI is still unclear, and the impact of circular RNAs (circRNAs) on AMI has rarely been recognized and needs to be explored.Materials and MethodsThe circRNA array was applied to investigate the expression level of circRNAs in the blood samples of coronary arteries of three AMI patients and three normal persons. Principal component analysis (PCA) and unsupervised clustering
doi:10.3389/fgene.2020.626492
pmid:33613625
pmcid:PMC7886988
fatcat:5y4bq2evi5drnlaovaedpjb4bi
more »
... s were performed to reveal the distinguished expression patterns of circRNAs. The miRNA expression profiles of AMI patients were identified from a public dataset from the Gene Expression Omnibus (GEO) database (GSE31568). The miRNA binding sites on the circRNAs were predicted by miRanda. The miRNA enrichment analysis and annotation tool were used to explore the pathways that the dysregulated circRNAs may participate in.ResultsIn total, 142 differentially expressed circRNAs, including 89 upregulated and 53 downregulated in AMI samples, were identified by the differential expression analysis. AMI patients had quite different circRNA expression profiles to those of normal controls. Functional characterization revealed that circRNAs that had the potential to regulate miRNAs were mainly involved in seven pathways, such as the Runt-related transcription factor-1 (RUNX1) expression and activity-related pathway. Specifically, hsa_circRNA_001654, hsa_circRNA_091761, hsa_circRNA_405624, and hsa_circRNA_406698 were predicted to sponge four miRNAs including hsa-miR-491-3p, hsa-miR-646, hsa-miR-603, and hsa-miR-922, thereby regulating RUNX1 expression or activity.ConclusionWe identified dysregulated blood circRNAs in the coronary arteries of AMI patients and predicted that four upregulated circRNAs were involved in the regulation of RUNX1 expression or activity through sponging four miRNAs.
DocStruct: A Multimodal Method to Extract Hierarchy Structure in Document for General Form Understanding
[article]
2020
arXiv
pre-print
., 2018; Liu et al., 2016; Wang and Hu, 2017) . Meanwhile, different directions have been proposed to deal with the form structure which is our focus in this paper. ...
arXiv:2010.11685v1
fatcat:mtbu67xqwbdyja3rugzzlj7sp4
Preservation of Cichoric Acid Antioxidant Properties Loaded in Heat Treated Lactoferrin Nanoparticles
2018
Molecules
In the current research, a new cichoric acid (CA) encapsulation system was investigated. The optimal condition for the formation of lactoferrin-cichoric acid nanoparticles (LF-CA NPs) was determined by controlling the solution pH, the thermal treatment conditions, and the concentration of CA. Fluorescence indicated that the electrostatic force and the hydrophobic force were the main forces in the formation of LF-CA NPs. LF-CA NPs prepared under different conditions were spherical in shape with
doi:10.3390/molecules23102678
fatcat:x6o2d6olfbditplkihghokdtza
more »
... maller particle sizes and good zeta potential demonstrating good colloidal stability. Especially, the prepared particle size of the LF-CA NPs at pH 7 and 95 °C was about 67.20 ± 1.86 nm. The circular dichroism (CD) and the Fourier transform infrared spectroscopy (FTIR) results showed that the combination of LF (lactoferrin) and CA affected the secondary structure of the LF. The differential scanning calorimetry (DSC) results indicated that the addition of CA increased the thermal stability of LF. In vitro antioxidant experiments confirmed the antioxidant capacity of LF-CA NPs was better than CA. CA was successfully encapsulated into LF NPs with high encapsulated efficiency (97.87–99.87%) by high performance liquid chromatography (HPLC). These results showed that LF could be used as the wall material of CA with excellent nature.
Identification and Characterization of Dipeptidyl Peptidase-IV Inhibitory Peptides from Oat Proteins
2022
Foods
In this study, flavourzyme, papain, neutrase, and alcalase, as well as gastrointestinal digestion simulated with pepsin and pancreatin, were used to hydrolyze oat protein, and the dipeptidyl peptidase-IV (DPP-IV) inhibitory activities of the oat protein hydrolysates were investigated. The results indicated that the oat protein hydrolysate by neutrase showed the most potent DPP-IV inhibitory property with an IC50 value of 2.55 ± 0.38 mg/mL. Using UPLC-MS/MS, ten new DPP-IV inhibitory peptides
doi:10.3390/foods11101406
pmid:35626976
pmcid:PMC9141920
fatcat:itdypiy6jzggll2gs5s2afhvce
more »
... e identified from the oat protein hydrolysate by neutrase. Among these peptides, IPQHY, VPQHY, VAVVPF, and VPLGGF exhibited the strongest DPP-IV inhibitory activity with IC50 values below 50 μM, and all of them acted as mixed-type inhibitors. Molecular docking indicated that the above four oat-derived peptides were predicted to form hydrogen bonds, attractive charge, and hydrophobic interactions with the residues of the active site of DPP-IV. Therefore, our results suggest that oat is an excellent protein source for food-derived DPP-IV inhibitory peptides and it has the prospect of becoming a dietary supplement for T2DM.
Multisource Data Analysis for Stock Prediction
2017
International Journal of u- and e- Service, Science and Technology
Over the last decade, quantitative investment has been adopted by the international investment community as a new method that combines financial data with mathematics and computer technology. This paper proposes a quantitative investment prediction model that combines multifactor and candlestick models to predict stock prices. First, relevant indicators with good future trends were identified to construct a stock pool containing stocks. Then, the candlestick model was used to decide when to buy
doi:10.14257/ijunesst.2017.10.7.02
fatcat:kpmkzxcqijhj5p2mqvocczo5uy
more »
... or sell stocks in the pool. The method developed here can be used for quantitative investment strategies, and the experimental results show it can have good returns in the investment stock market based on the transaction data. 12 Copyright ⓒ 2017 SERSC models have certain limitations. For example, they can determine the future trend for a specified period but cannot pinpoint a specific operating point in time; thus, there is no guarantee of profit [11] . Meanwhile, the candlestick chart, with its unique form of representation, was introduced into the analysis of stock and futures markets. This type of chart analysis is currently popular throughout Southeast Asia [12] [13] . Using a candlestick chart, we can completely record the daily or cyclical performance of a market in the form of special areas on a diagram, with different forms showing different meanings. From these morphological changes, we can infer some regular patterns [14] . Candlestick forms can be divided into the inversion of the form, the arrangement of the form, gap and trend lines, etc. From the candlestick, we can completely observe real market changes and see stockprice trends, while also grasping daily market fluctuations [15] [16] . In practice, stocks with good future trends should be selected in advance to guarantee high profits [17] . Each multifactor and candlestick method has advantages and disadvantages. Multifactor models can show stocks with good future trends but cannot tell the trader when to buy or sell, which is important for obtaining investment returns. While candlestick charts can signal when to buy or sell, the large number of stocks in the market makes it difficult for the trader to find the target stock using a candlestick chart. In practice, therefore, factor and candlestick methods can only be used for reference as technical analysis tools [18] . To overcome these drawbacks, we developed a model that combines multifactor and candlestick methods with moving average (MA), relative strength index (RSI), and other technical indicators to improve the stock index forecasting performance of the proposed model [19] [20] . Importantly, combining multifactor and candlestick models gives the model a high training speed, a highly effective learning algorithm, and structural simplicity [21] . Moreover, the combination model has the advantage of intelligent forecasting and compatible input and output for mapping stock data. When you want to illuminate which stocks to buy or sell, there is no requirement for prior information or characteristics. The combination model can find potentially profitable stocks and has low time complexity. In the existing literature, there is a knowledge gap regarding the identification of combinatory multifactor and candlestick models. The present study aims to fill that gap using the proposed model. The next section presents the multisource information fusion method. Then, Section 3 introduces the multifactor model, while Section 4 discusses the candlestick model. Finally, the results and conclusion are presented in Section 5 and 6, respectively. First, we constructed a matrix of candlestick chart patterns. The characteristics included candles, CCI, MA information, and rising and falling trends. To train the input and output sets, a support vector machine (SVM) machine learning model was used. Then, the SVM training was used to get the stock trading point.
Methionine Restriction Improves Gut Barrier Function by Reshaping Diurnal Rhythms of Inflammation-Related Microbes in Aged Mice
2021
Frontiers in Nutrition
*Correspondence: Xuebo Liu, xueboliu@nwsuaf.edu.cn ...
, Liu and Liu. ...
doi:10.3389/fnut.2021.746592
pmid:35004799
pmcid:PMC8733897
fatcat:y6vqufapu5fvbouwuk4xbedmp4
« Previous
Showing results 1 — 15 out of 321 results