Search

Your search keyword '"Wang, Chengxiao"' showing total 159 results
Start Over Author "Wang, Chengxiao"
159 results on '"Wang, Chengxiao"'

1. LATTE: Improving Latex Recognition for Tables and Formulae with Iterative Refinement

Author

Jiang, Nan, Liang, Shanchao, Wang, Chengxiao, Wang, Jiannan, and Tan, Lin

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

Portable Document Format (PDF) files are dominantly used for storing and disseminating scientific research, legal documents, and tax information. LaTeX is a popular application for creating PDF documents. Despite its advantages, LaTeX is not WYSWYG -- what you see is what you get, i.e., the LaTeX source and rendered PDF images look drastically different, especially for formulae and tables. This gap makes it hard to modify or export LaTeX sources for formulae and tables from PDF images, and existing work is still limited. First, prior work generates LaTeX sources in a single iteration and struggles with complex LaTeX formulae. Second, existing work mainly recognizes and extracts LaTeX sources for formulae; and is incapable or ineffective for tables. This paper proposes LATTE, the first iterative refinement framework for LaTeX recognition. Specifically, we propose delta-view as feedback, which compares and pinpoints the differences between a pair of rendered images of the extracted LaTeX source and the expected correct image. Such delta-view feedback enables our fault localization model to localize the faulty parts of the incorrect recognition more accurately and enables our LaTeX refinement model to repair the incorrect extraction more accurately. LATTE improves the LaTeX source extraction accuracy of both LaTeX formulae and tables, outperforming existing techniques as well as GPT-4V by at least 7.07% of exact match, with a success refinement rate of 46.08% (formula) and 25.51% (table).

Published
2024

2. Nova: Generative Language Models for Assembly Code with Hierarchical Attention and Contrastive Learning

Author

Jiang, Nan, Wang, Chengxiao, Liu, Kevin, Xu, Xiangzhe, Tan, Lin, Zhang, Xiangyu, and Babkin, Petr

Subjects
Computer Science - Software Engineering, Computer Science - Artificial Intelligence
Abstract

Binary code analysis is the foundation of crucial tasks in the security domain; thus building effective binary analysis techniques is more important than ever. Large language models (LLMs) although have brought impressive improvement to source code tasks, do not directly generalize to assembly code due to the unique challenges of assembly: (1) the low information density of assembly and (2) the diverse optimizations in assembly code. To overcome these challenges, this work proposes a hierarchical attention mechanism that builds attention summaries to capture the semantics more effectively and designs contrastive learning objectives to train LLMs to learn assembly optimization. Equipped with these techniques, this work develops Nova, a generative LLM for assembly code. Nova outperforms existing techniques on binary code decompilation by up to 14.84 -- 21.58% (absolute percentage point improvement) higher Pass@1 and Pass@10, and outperforms the latest binary code similarity detection techniques by up to 6.17% Recall@1, showing promising abilities on both assembly generation and understanding tasks.

Published
2023

21. Optimization of Extraction Process of Polysaccharides from Panax Notoginseng Residue and Its Kinetic Analysis.

Author

GU Liyun, LUO Zhiwei, WANG Chengxiao, and HAN Wei

Subjects
PANAX, POLYSACCHARIDES, RESPONSE surfaces (Statistics), HOT water, DIFFUSION coefficients, HEAT equation
Abstract

The optimization and kinetic analysis of ultrasound-assisted extraction of polysaccharides from Panax notoginseng residue was studied in this paper. On the basis of single factor experiment, the process was optimized by Box-Behnken Response Surface Methodology. The optimal extraction process was as follows: ultrasound temperature 60 °C, ultrasound power 136.8 W, ultrasound time 53 min, liquid-solid ratio 43 mL/g. Under the optimal conditions, the yield of polysaccharide was 6.79%, while the predicted yield was 6.88% and the relative standard deviation (RSD) was 1.31%. In comparison with the same conditions of hot water extraction, the yield was increased by more than 55% by the optimization. The kinetic model of ultrasound-assisted extraction of polysaccharides from Panax notoginseng residue was built. The rate constant and relative extraction residual rate of polysaccharides was calculated by measuring the concentration of polysaccharides in the extraction solution at different ultrasound powers and different ultrasound times. According to those two parameters, the half-life regression equation and the effective diffusion coefficient regression equation were fitted. The experimental results showed that the rate constant increased with the increasing of ultrasonic power. The enhancement of ultrasonic power was beneficial to the faster dissolution of polysaccharides from Panax notoginseng residue. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

23. Nova$^+$: Generative Language Models for Binaries

Author

Jiang, Nan, Wang, Chengxiao, Liu, Kevin, Xu, Xiangzhe, Tan, Lin, Zhang, Xiangyu, Jiang, Nan, Wang, Chengxiao, Liu, Kevin, Xu, Xiangzhe, Tan, Lin, and Zhang, Xiangyu

Abstract

Generative large language models (LLMs) pre-trained on code have shown impressive effectiveness in code generation, program repair, and document analysis. However, existing generative LLMs focus on source code and are not specialized for binaries. There are three main challenges for LLMs to model and learn binary code: hex-decimal values, complex global dependencies, and compiler optimization levels. To bring the benefit of LLMs to the binary domain, we develop Nova and Nova$^+$, which are LLMs pre-trained on binary corpora. Nova is pre-trained with the standard language modeling task, showing significantly better capability on five benchmarks for three downstream tasks: binary code similarity detection (BCSD), binary code translation (BCT), and binary code recovery (BCR), over GPT-3.5 and other existing techniques. We build Nova$^+$ to further boost Nova using two new pre-training tasks, i.e., optimization generation and optimization level prediction, which are designed to learn binary optimization and align equivalent binaries. Nova$^+$ shows overall the best performance for all three downstream tasks on five benchmarks, demonstrating the contributions of the new pre-training tasks.

Published
2023

24. Cell Membrane-Derived Nanovehicles for Targeted Therapy of Ischemic Stroke: From Construction to Application.

Author

Hao, Cui, Sha, Ma, Ye, Yang, and Wang, Chengxiao

Subjects
ISCHEMIC stroke, DRUG delivery systems, STROKE, CANCER cells, DRUG dosage
Abstract

Ischemic stroke (IS) is a prevalent form of stroke and a leading cause of mortality and disability. Recently, cell membrane-derived nanovehicles (CMNVs) derived from erythrocytes, thrombocytes, neutrophils, macrophages, neural stem cells, and cancer cells have shown great promise as drug delivery systems for IS treatment. By precisely controlling drug release rates and targeting specific sites in the brain, CMNVs enable the reduction in drug dosage and minimization of side effects, thus significantly enhancing therapeutic strategies and approaches for IS. While there are some reviews regarding the applications of CMNVs in the treatment of IS, there has been limited attention given to important aspects such as carrier construction, structural design, and functional modification. Therefore, this review aims to address these key issues in CMNVs preparation, structural composition, modification, and other relevant aspects, with a specific focus on targeted therapy for IS. Finally, the challenges and prospects in this field are discussed. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

25. Applications for natural deep eutectic solvents in Chinese herbal medicines

Author

Li, Minghui, Rao, Cheng, Ye, Xiaoqian, Wang, Mei, Yang, Boyuan, Wang, Chengxiao, Guo, Liqun, Xiong, Yin, and Cui, Xiuming

Subjects
Pharmacology, Pharmacology (medical)
Abstract

Chinese herbal medicines (CHMs), with a wide range of bioactive components, are considered to be an important source for new drug discovery. However, the process to isolate and obtain those bioactive components to develop new drugs always consumes a large amount of organic solvents with high toxicity and non-biodegradability. Natural deep eutectic solvents (NADES), a new type of green and designable solvents composed of primary plant-based metabolites, have been used as eco-friendly substitutes for traditional organic solvents in various fields. Due to the advantages of easy preparation, low production cost, low toxicity, and eco-friendliness, NADES have been also applied as extraction solvents, media, and drug delivery agents in CHMs in recent years. Besides, the special properties of NADES have been contributed to elucidating the traditional processing (also named Paozhi in Chinese) theory of CHMs, especially processing with honey. In this paper, the development process, preparation, classification, and applications for NADES in CHMs have been reviewed. Prospects in the future applications and challenges have been discussed to better understand the possibilities of the new solvents in the drug development and other uses of CHMs.

Published
2023
Full Text
View/download PDF

37. Salicylic acid reduces the accumulation of aluminum in Panax notoginsen root cell wall pectin via the NO signaling pathway

Author

Yang Xiaoyan, Qu Yuan, Wang Chengxiao, Dai Chunyan, Cui Xiuming, Yang Ye, Guo Lanping, and Liu Diqiu

Subjects
0106 biological sciences, 0301 basic medicine, food.ingredient, biology, Pectin, food and beverages, Soil Science, Endogeny, Plant Science, Phenylalanine ammonia-lyase, biology.organism_classification, 01 natural sciences, Nitric oxide, Pectinesterase, Cell wall, 03 medical and health sciences, chemistry.chemical_compound, 030104 developmental biology, food, chemistry, Biochemistry, Panax notoginseng, Salicylic acid, 010606 plant biology & botany
Abstract

Salicylic acid (SA) and nitric oxide (NO) are key signal molecules involved in the reduction of Al accumulation in many plants. The purpose of this study was to investigate whether they could reduce the root content of Al and/or its binding to the root cell wall (CW) fractions in Panax notoginseng, and to investigate the corresponding regulatory mechanisms. Effects of Al treatments on the endogenous SA and NO contents of P. notoginseng were detected. Exogenous SA and paclobutrazol (PAC, a SA synthesis inhibitor) treatments were conducted to test the effects on endogenous NO content; exogenous sodium nitroprusside (SNP, a NO doner) and 2-phenyl-4,4,5,5-tetramethyl- imidazoline-1-oxyl-3-oxyde (cPTIO, a NO synthesis inhibitor) treatments were conducted to test the effects on endogenous SA content. Pectin methyltransferase (PMT) and pectin methylesterase (PME) activity and the gene expression, pectin methylesterification degree (PMD), pectin content, and the accumulation of Al in roots were also studied under the treatments described above. Al stress induced a significant increase of the activity and expression of phenylalanine ammonia-lyase (PAL), and promoted the significant increase of endogenous SA content of P. notoginseng roots. The Al promoted accumulation of endogenous NO could be enhanced by exogenous SA treatment, but was reduced by PAC. Accumulation of endogenous NO that promoted by Al could be enhanced by exogenous SA treatment, but was reduced by PAC treatment. However, SNP or cPTIO treatments had no significant effect on the Al-induced endogenous SA content of P. notoginseng. This showed that NO acted downstream of SA signaling of P. notoginseng under Al stress. Al stress can increase the activity and genes expression of PME in the roots of P. notoginseng, reduce PMD, increase CW pectin content, and thus, the binding capacity of CW pectin to Al was enhanced in P. notoginseng roots. Exogenous SA or SNP both reduced the Al-binding capacity of root CW pectin by decreasing the pectin content, and increased the PMD by inhibiting the genes expression and activity of PME. The effects of PAC or cPTIO on the above-mentioned indicators under Al stress were opposite to that of exogenous SA or SNP treatments. The Al stress induced accumulation of pectin and the reduction of PMD in the root CW of P. notoginseng could be reversed by the treatments of exogenous SA or SNP, then the Al contents in root CW pectin were reduced. Al stress activated the endogenous SA and NO signaling pathways in P. notoginseng, and NO acted downstream of SA. It showed that the Al-activated NO-SA signaling pathway contributed in the reduction of Al binding to the root under Al stress.

Published
2018
Full Text
View/download PDF

40. Distribution pattern of aluminum in Panax notoginseng, a native medicinal plant adapted to acidic red soils

Author

Qu Yuan, Wang Chengxiao, Yang Xiaoyan, Chen Qi, Yang Ye, Dai Chunyan, Guo Lanping, Cui Xiuming, and Liu Diqiu

Subjects
0106 biological sciences, 0301 basic medicine, food.ingredient, biology, Pectin, Chemistry, Soil Science, Plant Science, biology.organism_classification, Hydroponics, 01 natural sciences, Pectinesterase, Rhizome, Cell wall, 03 medical and health sciences, Horticulture, 030104 developmental biology, food, Soil pH, Panax notoginseng, Red soil, 010606 plant biology & botany
Abstract

Panax notoginseng is normally cultivated in acidic red soil with high soluble Al, however, the distribution pattern of Al in P. notoginseng is unknown. The aim of this study is to investigate the Al distribution in different organs of P. notoginseng grown in different acid soil, in order to get a deeper insight into the Al tolerance mechanism in this species. P. notoginseng was sampled from different sites and subjected to Al determination. Al tolerance and accumulation were investigated by exposing the seedlings to different Al concentrations in hydroponic solution. The expression of pectin methylesterase genes was determined by quantitative RT-PCR. The soil pH ranged from 4.68 to 6.13. The total Al concentration in soil was 76.7–168.5 g·kg−1 and active Al was 1.34–5.01 g·kg−1. Al concentration differed with organs of P. notoginseng, following the order: rootlet > rhizome > main root > leaf and stem. Furthermore, Al concentration in the cell wall, organelle and cytoplasmic supernatant of P. notoginseng was approximately 15:2:1 and most Al in the cell wall was bound to pectin component. Al treatment significantly increased the content of pectin and enhanced the expression of PME3, PME7, and PME40 of P. notoginseng root. P. notoginseng was able to grow under high active Al condition and accumulated high Al in its underground parts. The pectin component in the cell wall was the major binding site of Al of P. notoginseng root.

Published
2017
Full Text
View/download PDF

45. A study of the freeze-drying process and quality evaluation of Angelica sinensis.

Author

Haizhu, Zhang, Zheng, Luyao, Zhang, Xingying, Cui, Xiuming, Wang, Chengxiao, and Qu, Yuan

Subjects
DONG quai, FERULIC acid, FREEZE-drying, DRYING
Abstract

The freeze-drying process of Angelica sinensis (Oliv.) Diels was studied and evaluated. Using a single factor investigation, drying temperature and pressure were determined as the main factors affecting the drying process. The central composite design (CCD) combined with response surface method was employed to optimize the drying process. Optimal conditions were determined to be 52 Pa, 63 °C, and a slice thickness of 5 mm. Subsequently, samples were compared in terms of chemical constituents, microstructure, and in vitro absorption profiles under different drying operations. The freeze-drying process was effective for the preservation of ferulic acid (1.82 mg/g), Z-ligustilide (13.91 mg/g), and other compositions. The porous and loose characteristic structure enabled rapid release of ferulic acid (71%, 60 min) and Z-ligustilide (32%, 60 min). Therefore, the freeze-drying method is a reasonable and efficient drying method for the dehydration of A. sinensis. [ABSTRACT FROM AUTHOR]

Published
2021
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results