17,507 results on '"Huang Jun"'
Search Results
2. Nursing of integrated traditional Chinese and Western medicine in a case of severe diabetic ketoacidosis complicated with intestinal obstruction and pneumonia caused by immunotherapy for gastric cancer (1例胃癌免疫治疗引起重度糖尿病酮症酸中毒伴肠梗阻并发肺炎的中西医结合护理体会)
- Author
-
LIU Yin (柳吟), HUANG Jun (黄君), QIU Minyan (邱敏燕), ZHAO Hongxia (赵红霞), and SHEN Xiaoyan (沈晓燕)
- Subjects
gastric cancer ,immune checkpoint inhibitors ,diabetes ketoacidosis ,intestinal obstruction ,integrated traditional chinese and western medicine nursing ,chemotherapy ,胃癌 ,免疫检查点抑制剂 ,糖尿病酮症酸中毒 ,肠梗阻 ,中西医结合护理 ,化疗 ,Nursing ,RT1-120 - Abstract
This paper summarized the integrated traditional Chinese and Western medicine nursing of a case of severe diabetic ketoacidosis (DKA) with acute intestinal obstruction complicated with pneumonia secondary to gastric adenocarcinoma treated with anti-PD-1. The main points of nursing included: disease monitoring, nausea and vomiting nursing, lack of knowledge and other general nursing measures. The key and difficult points of nursing for this case were complicated with intestinal obstruction and ICPis related diabetes. Therefore, integrated traditional Chinese and western medicine nursing and corresponding nursing measures were taken to relieve the clinical symptoms and improve the rehabilitation of the patient. (本文总结1例基于胃癌免疫治疗引起重度糖尿病酮症酸中毒(DKA)伴肠梗阻并发肺炎患者的中西医结合护理经验。患者病情复杂, 合并症多, 在加强病情监测、心理护理、恶心呕吐护理、健康宣教等常规护理的基础上, 针对并发症采取肠梗阻护理干预、DKA急性发作期护理干预, 改善患者症状, 促进患者早期康复。)
- Published
- 2024
- Full Text
- View/download PDF
3. Analysis of energy conservation potential of water cold storage in data centers in Pearl River Delta
- Author
-
DONG Zijun, HUANG Jun, SHAO Shuangquan, YANG Minghong, and WANG Bo
- Subjects
data center ,refrigeration system ,water cold storage ,economic analysis ,Materials of engineering and construction. Mechanics of materials ,TA401-492 ,Environmental engineering ,TA170-171 - Abstract
In order to reduce the energy consumption of data centers in the southern coastal region and enhance the analysis of the water cold storage system, the water cold storage system was selected as the energy storage solution based on the cooling load of data centers in the Pearl River Delta region and the local time-of-use tariff policy. First, the water cold storage system and operation control scheme were optimized; second, the operation situation, energy consumption and economy of the water cold storage system were analyzed for typical days, months and whole year. The results show that the optimized water cold storage system can guarantee the indoor end-load demand and operate efficiently, and the annual cooling operation cost is reduced by 25.78%, which has a good effect of "peak-load shifting effect" and reduces the cooling electricity cost significantly.
- Published
- 2023
- Full Text
- View/download PDF
4. Countermeasures for Low-Carbon Transformation of Construction Industry in China Toward the Carbon Peaking and Carbon Neutrality Goals
- Author
-
Wu Zezhou, Huang Haoquan, Chen Xiangsheng, Li Jianjun, He Qiufeng, Li Ao, Huang Jun, Lin Yuhan, Liu Xingyu, and Wang Jiahao
- Subjects
carbon peaking and carbon neutrality goals ,construction industry ,carbon emissions ,low-carbon transformation ,Engineering (General). Civil engineering (General) ,TA1-2040 - Abstract
Carbon emission reduction in the construction industry is crucial for promoting low-carbon transformation of the industry and for realizing the carbon peaking and carbon neutrality goals. This study analyzes the current status of carbon emissions in Chinese construction industry. It identifies that the production of building materials and the operation of buildings are the primary sources of carbon emissions in the overall construction process. Moreover, the growth rate of carbon emissions from the construction industry in China is decelerating, and the spatial characteristics of carbon emissions demonstrate an increase from the south to the north and from the west to the east. We have identified six key challenges that must be addressed to achieve carbon peaking and carbon neutrality in the construction industry of China: policy and standards framework, energy-efficient and carbon-reducing technologies, reduction of carbon emissions from existing buildings, carbon emission data collection and monitoring systems, economic incentives, and increased awareness of carbon reduction. Staged goals for the low-carbon development of the construction industry are also proposed. Furthermore, we suggest improving related policies and standards system, optimizing the use of energy-saving technologies, reducing carbon emissions in existing buildings, developing green finance, improving the ability to monitor carbon emissions, and enhancing the awareness of energy conservation and carbon reduction, so as to provide references for the low-carbon transformation and high-quality development of China’s construction industry.
- Published
- 2023
- Full Text
- View/download PDF
5. Testing and Evaluation of Hard Sandstone Aggregate in Hot Mix Asphalt
- Author
-
Liu PeiRong, Yang LiMing, Huang Jun, Li ChaoYang, and Yang DaTian
- Subjects
Materials of engineering and construction. Mechanics of materials ,TA401-492 - Abstract
To address the shortage of the aggregate used in a hot mix asphalt (HMA) pavement in Guangxi, properties such as the aggregate crushing, polished stone, and Los Angeles abrasion values of a type of hard sandstone aggregate used in HMA were tested after various conditioning treatments. The hard sandstone aggregate met the technical requirements for aggregate in HMA. In addition, the influence of the Marshall compaction on the hard sandstone aggregate-combined grading was tested. The combined grading curve changed a little, and the aggregate satisfied the corresponding technical requirements. Therefore, according to the abovementioned results, the hard sandstone aggregate can be used as a coarse aggregate in HMA.
- Published
- 2024
- Full Text
- View/download PDF
6. Establishment and application of a zirconium alloy irradiation deformation data mining database based on a mesoscale model
- Author
-
HUANG Jun, ZHANG Zikang, LIU Guisen, and SHEN Yao
- Subjects
zirconium alloy ,irradiation deformation ,mesoscale models ,data mining database ,visco plastic self consistent model ,Nuclear engineering. Atomic power ,TK9001-9401 - Abstract
BackgroundIn the nuclear industry, zirconium alloys, employed as structural materials, undergo irradiation-induced deformation owing to the combined effects of thermo, mechanical stress, and neutron irradiation during the irradiation process, thereby affecting their reliability in use.PurposeThis study aims to explore and predict the irradiation deformation of zirconium alloys to ensure the safety and economical operation of reactors.MethodsFirst of all, a database of experimental data on irradiation deformation of zirconium alloys from publicly available literature was developed by collection, cleaning, re-organization and placement of these data into a well-designed database, accompanied by detailed descriptions of these data. Then, data patterns, model validation and data mining were conducted by combining the database with our own developed mesoscopic model. Specifically, over fifty sets of preliminary data from the data mining library were subjected to data mining to analyze the correlation between model parameters and control variables.ResultsData mining results indicate that the critical resolved shear stress (CRSS) of the slip system increases with Nb content whilst the irradiation creep compliance of zirconium alloys increases with temperature.ConclusionsCompared with conventional data induction methods, this study introduces a novel approach for fitting and data mining using a physics-based mesoscopic model, offering fresh perspectives and methodologies for studying irradiation-induced deformation in zirconium alloys. The impact of this approach will be boosted as the collected dataset expands.
- Published
- 2024
- Full Text
- View/download PDF
7. Rare presentation of double-clonal Waldenström macroglobulinemia with pulmonary embolism: A case report
- Author
-
Sun Yan, Meng Fan-Jun, Huang Jun-Xia, Yan Xue-Shen, Zhao Xia, Zhou Jing-Jing, and Gao Yan
- Subjects
case report ,double-clone ,lymphoplasmacytic lymphoma ,pulmonary embolism ,waldenström macroglobulinemia ,Biology (General) ,QH301-705.5 - Abstract
Waldenström macroglobulinemia (WM) rarely leads to pulmonary embolism. Due to its low incidence, the underlying pathophysiology, prognosis, and optimal treatment remain largely unexplored and uninvestigated. In this study, a patient with a double-clonal WM, a rare subtype, presented with pulmonary embolism. The patient had a small number of plasma cells without morphological abnormalities, and an effective therapeutic response was observed. Nonetheless, the clinical prognosis requires a long-term follow-up.
- Published
- 2023
- Full Text
- View/download PDF
8. The County Network Structure and Driving Mechanism of China's Cold Chain Logistics Enterprises from a Multi-Scale Perspective
- Author
-
Li Xiaomei and Huang Jun
- Subjects
multi-scale ,cold chain logistics enterprises ,county network ,driving mechanism ,cold chain logistics corridors ,Geography (General) ,G1-922 - Abstract
To expand micro-scale research on the urban network of cold chain logistics enterprises, enrich cold chain logistics channel theories, guide marginal counties to improve their network status, further promote the rational allocation of cold chain logistics resources at various scales, and promote the coordinated development of regional cold chain logistics, this study constructed three levels of county network respectively from the inside of the cold chain logistics corridors,within and beyond the corridors and at a national scale. Using Social Network Analysis, Motif analysis, geodetectors, and other models, it gradually analyzed the county network structure of cold chain logistics enterprises in 2020 from the overall to the local level and discussed the driving mechanism of the network structure. The results show that: 1) from an overall perspective, the network structure at all scales presents a core-periphery layout, and the core counties with archipelagic distribution are mainly located in Beijing Tianjin Hebei, Yangtze River Delta, and Pearl River Delta. The difference is that network development inside and outside the corridors and at the national scale is not yet mature, and the network status and node relevance of counties at various scales are quite different; 2) from a local perspective, the core counties within the corridors mainly establish a dual relationship with the core counties in the same city, build a ternary closed and mutually beneficial group relationship with other counties, and develop multi-corridor multi-core communities. The outer core counties of the inner and outer scales of the corridors are mostly formed near the hub of the corridor, often occupying the structural hole position of the inner and outer links of the corridors and are closely linked with other marginal counties in the main agricultural production areas. The core counties at the national scale radiate mainly to the outside world. They often choose the eastern coastal open cities or ports as windows to absorb multiple investments, leading to economic circles and economic belt associations; and 3) from the perspective of the core-periphery layout driving mechanism, economic factors, such as consumption level, are the common driving force of all scales; spatial factors, such as geographical location, lead to differences between scales; and aviation facilities do not show an obvious driving force. The theory of urban network driving mechanisms is still applicable to the study of county units. The formation of the county network structure stems from the functional differentiation and locational separation behavior of cold chain logistics enterprises. Driving mechanisms, such as scale preference, endowment preference, and network proximity effect, show heterogeneity in the role of network structures at various scales.
- Published
- 2023
- Full Text
- View/download PDF
9. Retraction of 'Two mixed-ligand coordination polymers based on 2,5-thiophenedicarboxylic acid and flexible N-donor ligands: The protective effect on periodontitis via reducing the release of IL-1β and TNF-α'
- Author
-
Wu Shao-Hsuan and Huang Jun-Hui
- Subjects
Chemistry ,QD1-999 - Published
- 2023
- Full Text
- View/download PDF
10. Effects of different feeding patterns on physical and nutritional status of infants aged 6‒12 months
- Author
-
JIANG Chunhua, HUANG Jun, LI Yun, ZHANG Ying, KANG Shurong, ZHANG Jing, LI Wenxian, JIANG Hong, and XU Xiaoxi
- Subjects
feeding pattern ,physical growth ,nutritional status ,infant ,Medicine - Abstract
ObjectiveTo analyze the effects of different feeding patterns on the physical and nutritional status of children aged 6‒12 months, so as to provide reference for promoting scientific feeding and health development of infants and young children.MethodsChildren born between December 2019 and February 2020 and who had completed three follow-up visits at 6‒, 9‒ (8‒10 months) and 12‒ (11‒14 months) months old in all of the 13 communities of Minhang, Shanghai were selected. The subjects’ basic information was investigated by questionnaires. The indicators including feeding pattern, physical development (body weight, body length, head circumference) and nutritional status (the detection rate of overweight, obesity, low body weight, growth retardation, emaciation and iron deficiency anemia) were followed up in the outpatient department, with iron deficiency anemia only monitored at the 6‒ and 12‒ months old. According to different feeding patterns, the groups of 6‒ months old were divided into three groups of exclusive breast feeding (EBF), mixed feeding (MF) and artificial feeding (AF), while 9‒ and 12‒ months old were divided into MF and AF groups. The differences of basic information and follow-up results among the groups were analyzed.ResultsA total of 470 children were included, including 130 (27.66%), 288 (61.28%) and 52 (11.06%) respectively in EBF, MF and AF groups at the 6‒ months old,and 319 (67.87%) and 196 (41.70%) in MF group at the 9‒ and 12‒ months old. There was no significant difference in the other follow-up results among the groups. The detection rate of iron deficiency anemia in 6‒ months old EBF (13.08%) was higher than that in MF group (5.90%) and AF group (1.92%) (χ2=8.40, P=0.010), while it was still higher in 12‒ months old MF group (9.69%) than in AF group (2.92%) (χ2=9.68, P=0.002).ConclusionThere is no significant difference in body weight,body length, head circumference, and the detection rates of overweight, obesity, low body weight, growth retardation and emaciation among the groups of different feeding patterns in the children aged 6‒12 months. The detection rate of iron deficiency anemia in the EBF and MF groups is significantly higher than that in the AF groups of children aged 6‒ and 12‒ months old.
- Published
- 2023
- Full Text
- View/download PDF
11. Research on the Influence of Model Surface Conductivity on RCS Test Accuracy
- Author
-
Wu Yacong, Huang Jun, and Song Lei
- Subjects
Surface resistance ,RCS test model ,quantitative analysis ,Chinese military standard ,accuracy requirement ,Electrical engineering. Electronics. Nuclear engineering ,TK1-9971 - Abstract
The accuracy of the RCS (Radar Cross Section) test is closely related to the conductivity of the model surface. The Chinese military standard has stringent surface resistance criteria for models. If the resistance criteria are appropriately relaxed, model processing can be more effective and expenses can be reduced. However, there are few quantitative evaluations of the relationship between surface resistance and RCS test error, making it impossible to design the model surface resistance in accordance with RCS test accuracy requirements. In this paper, firstly, the influence of model surface resistance on RCS is clarified through theoretical, simulation, and experimental research. The RCS analysis approach of the finite conductivity model is established. Subsequently, taking various typical shapes as research objects, such as the sphere, cube, flying wing aircraft, wing-body combination, and stealth aircraft, the simulation analysis of radar scattering characteristics is carried out under varied sheet resistances and multiple frequencies. A more attainable surface conductivity control standard for the RCS test model has been formulated, which is relaxed by 93% and 727% for models with average RCS of less than and higher than −30dBsm compared with the indicator in the Chinese military standard.
- Published
- 2023
- Full Text
- View/download PDF
12. Study on Fatigue Crack Propagation Rate and Corrosion Mechanism of 60Si2Mn Spring Steel under Salt Spray Environment
- Author
-
HUANG Jun, WU Ceng, MA Wei, YU Jian
- Subjects
60si2mn ,spring steel ,salt spray test ,corrosion rate ,crack propagation rate ,corrosion mechanism ,Materials of engineering and construction. Mechanics of materials ,TA401-492 ,Technology - Abstract
With the aim of studying the fatigue crack propagation rate law and corrosion mechanism of 60Si2Mn, a commonly used material for high-speed track ballastless railway fastener clips in coastal environment,a neutral salt spray chamber was used to simulate the marine environment. The fatigue evaluation of 60Si2Mn spring steel compact tensile (CT) samples under different salt spray corrosion time conditions was carried out and the corrosion mechanism was analyzed through salt spray corrosion and fatigue crack propagation experiments. The surface morphology of the corroded samples was observed with a scanning electron microscope (SEM),and the corrosion products of the samples were analyzed with an X-ray diffractometer (XRD). Then the fatigue crack propagation rate test was carried out, the crack propagation length was measured based on the flexibility method, and the test data was processed by the seven-point incremental polynomial method. Results showed that with the increase of corrosion time, the corrosion rate of the sample had a downward trend, and the data points of its mass loss versus corrosion time followed the power function law. Compared with the uncorroded sample,the fatigue crack propagation rate of the material did not change significantly. With the increase of the corrosion time, the surface of the sample was covered by a rust layer composed of Fe3O4 and FeOOH, which effectively prevented the mediums such as Cl- from continuing to erode the substrate, making the material possess good corrosion resistance.
- Published
- 2022
- Full Text
- View/download PDF
13. Determination of azelaic acid and potassium azeloycinate diglycinate in cosmetics by HPLC
- Author
-
XU Yong, HAN Jing, YU Ling, HUI Taicheng, HUANG Jun, PENG Xingsheng, and ZHENG Rong
- Subjects
azelaic acid ,potassium azeloycinate diglycinate ,hplc ,cosmetics ,Medicine - Abstract
ObjectiveAn HPLC method was established for the determination of azelaic acid and potassium azeloycinate diglycinate in cosmetics.MethodsThe samples were extracted with 60 mmol·L-1 sodium hydroxide water solution-methyl alcohol. After centrifugation and filtration, the analysis of azelaic acid and potassium azeloycinate diglycinate was performed with a SVEA C8(250 mm×4.6 mm, 5 μm) column, using 15 mmol‧L-1 potassium dihydrogen phosphate solution (pH=3.0) and acetonitrile for gradient elution at a flow rate of 1.0 mL·min-1. The analytes were detected with UV detector, and quantified by external standard curve.ResultsThe results showed a good linearity in the range of 5‒1 000 μg‧mL-1 with correlation coefficients (r) larger than 0.999. The detection limit of azelaic acid and potassium azeloycinate diglycinate (LOD) was 0.020% and 0.015%, respectively. The spiked recoveries were 87.66% to 108.96% with the relative standard deviation (RSD) of 0.6% to 3.3%.ConclusionThe method is simple, rapid and highly sensitive. It is suitable for the determination of azelaic acid and potassium azeloycinate diglycinate in cosmetics.
- Published
- 2022
- Full Text
- View/download PDF
14. Health literacy of child rearing and its determinants among caregivers of 0‒3 year-old children in Minhang District, Shanghai
- Author
-
CHEN Min, HUANG Jun, ZHANG Yue, ZHANG Ying, JIANG Chunhua, XUE Lili, and LI Yun
- Subjects
child rearing ,health literacy ,determinant ,shanghai ,Medicine - Abstract
ObjectiveTo investigate the status of health literacy of child rearing among caregivers of 0‒3 year-old children and analyze its determinants in Minhang District, Shanghai to provide scientific evidence for further promoting health literacy of caregivers.MethodsThe caregivers of 0‒3 year-old children were randomly selected and investigated using a questionnaire of health literacy in 13 community health care centers of Minhang District from March 2020 to April 2021.ResultsA total of 2819 valid questionnaires were collected. The total score of health literacy of child rearing for caregivers were 159.2±23.1 (full score being 200.0). The score of physical health literacy and mental health literacy were 80.0±10.2 and 79.1±15.1, respectively (full score being 100.0). The proportions of health literacy in the caregivers were: scarce 3.1%, insufficient 7.6%, good 47.2%, and excellent 42.1%. The multivariate ordinal Logistic regression model showed the following were independent determinants: always (OR=5.90) or often (OR=3.55) choosing information from professional approach like hospital to guide children rearing,mother (OR=4.98) or father (OR=3.95) being caregivers, easy access to children rearing information (OR=4.50), mother education degree being master and above (OR=3.10) or junior college, undergraduate (OR=2.14), father education degree being junior college, undergraduate (OR=1.59) or master and above (OR=1.53), single child (OR=1.37), household registration being Shanghai (OR=1.37). All caregivers with the above factors showed more superior health literacy.ConclusionThe level of health literacy of child rearing among caregivers of 0‒3 year-old children in Minhang District has a good level of health literacy. Targeted health education should be strengthened for caregivers of children in the future.
- Published
- 2022
- Full Text
- View/download PDF
15. Efficacy of intravenous thrombolysis combined with mechanical stent interventional thrombectomy on acute ischemic stroke
- Author
-
Huang Jun, Zhang Ming, Nie Qingbin, Zhang Xinye, He Xin, Yang Yufeng, and Mao Gengsheng
- Subjects
thrombolysis ,mechanical thrombectomy ,acute ischemic stroke ,efficacy ,Biochemistry ,QD415-436 - Abstract
Background: To investigate the efficacy and safety of intravenous thrombolysis combined with mechanical stent interventional thrombectomy in the treatment of acute ischemic stroke. Methods: A retrospective analysis was carried out for clinical data of 118 patients with acute ischemic stroke. The patients enrolled were divided into control group (recombinant tissue plasminogen activator (rt-PA) intravenous thrombolysis) and thrombectomy group (mechanical stent interventional thrombectomy based on rt-PA intravenous thrombolysis). The vascular recanalization rate and clinical efficacy after treatment were compared between the two groups. National Institutes of Health Stroke Scale (NIHSS) was used to identify the degree of neurological impairment in all patients before and after treatment, and Barthel Index was used to assess their activity of daily living. Moreover, the changes in the levels of T-lymphocyte subpopulation in peripheral blood and immuno-inflammatory factors before and after treatment were compared, and prognosis of patients and incidence of adverse reactions were recorded. Results: The response rate inthrombectomy group (93.2%) was significantly better than that in control group (76.3%). The NIHSS sore and modified Rankin scale (mRS) score after treatment were significantly lower than those before treatment, while the Barthel Index after treatment was distinctly higher than that before treatment. The NIHSS score and mRS score in thrombectomy group obviously declined compared with those in control group at 1 month after treatment. The Barthel Index in thrombectomy group was obviously higher than that in control group at 1 month and 2 months after treatment. Levels of cluster of differentiation 3 (CD3)+, CD3+CD4+, CD4+/CD8+ and natural killer (NK) cells in peripheral blood at 6 months after treatment evidently rose compared with those before treatment, while level of CD3+CD8+ evidently declined compared with that before treatment. In thrombectomy group, levels of CD3+, CD3+CD4+, CD4+/CD8+ and NK cells were markedly higher than those in control group, while the level of CD3+CD8+ was markedly lower than that in control group. Besides, in thrombectomy group, levels of serum osteopontin (OPN), malondialdehyde (MDA) and N-terminal pro-B-type natriuretic peptide (NT-proBNP) were evidently lower than those in control group at 1 month after treatment, while the level of serum superoxide dismutase (SOD) was evidently higher than that in control group. Compared with that in control group, the acute vascular reocclusion rate in thrombectomy group was significantly decreased at 3 months after treatment (10.2% vs. 22.0%). Conclusions: Intravenous thrombolysis combined with mechanical stent interventional thrombectomy can effectively promote the vascular recanalization, improve the neurological function and activity of daily living of patients, reinforce the immunological function, inhibit the oxidative stress response and improve the prognosis of patients.
- Published
- 2022
- Full Text
- View/download PDF
16. Cathepsin a upregulation in glioma: A potential therapeutic target associated with immune infiltration
- Author
-
Zhang Ming, Huang Jun, Wang Yunfei, Nie Qingbin, Zhang Xinye, Yang Yufeng, and Mao Gengsheng
- Subjects
cathepsin a ,glioma ,prognostic biomarker ,immune infiltration ,Biochemistry ,QD415-436 - Abstract
Background: Glioma is the result of malignant transformation of glial cells in the white matter of the brain or spinal cord and accounts for approximately 80% of all intracranial malignancies. Cathepsin A (CTSA) is highly expressed in a variety of tumor tissues, but its role in glioma is poorly studied. This study analyses the relationship between CTSA, and glioma based on The Cancer Genome Atlas (TCGA). Methods: Data for glioma patients were collected from TCGA. The expression level of CTSA was compared between paired glioma tissues and normal tissues with Wilcoxon rank-sum test. In addition, the Wilcoxon ranksum test was also applied to analyze the relationship between clinicopathologic features and CTSA expression. Kaplan-Meier Plotter was applied to analyze OS, DSS and PFI. Immuno-infiltration analysis of BLCA was performed by single sample gene set enrichment analysis (ssGSEA) in the "GSVA" R package. Results: The CTSA was overexpressed in glioma tissues compared to normal tissues (P
- Published
- 2022
- Full Text
- View/download PDF
17. DocLayLLM: An Efficient and Effective Multi-modal Extension of Large Language Models for Text-rich Document Understanding
- Author
-
Liao, Wenhui, Wang, Jiapeng, Li, Hongliang, Wang, Chengyu, Huang, Jun, and Jin, Lianwen
- Subjects
Computer Science - Computer Vision and Pattern Recognition - Abstract
Text-rich document understanding (TDU) refers to analyzing and comprehending documents containing substantial textual content. With the rapid evolution of large language models (LLMs), they have been widely leveraged for TDU due to their remarkable versatility and generalization. In this paper, we introduce DocLayLLM, an efficient and effective multi-modal extension of LLMs specifically designed for TDU. By integrating visual patch tokens and 2D positional tokens into LLMs and encoding the document content using the LLMs themselves, we fully take advantage of the document comprehension capability of LLMs and enhance their perception of OCR information. We have also deeply considered the role of the chain-of-thought (CoT) and innovatively proposed the techniques of CoT Pre-training and CoT Annealing. Our DocLayLLM can achieve remarkable performances with lightweight training settings, showcasing its efficiency and effectiveness. Experimental results demonstrate that our DocLayLLM surpasses existing OCR-dependent methods and also outperforms OCR-free competitors.
- Published
- 2024
18. Rotation center identification based on geometric relationships for rotary motion deblurring
- Author
-
Qin, Jinhui, Ma, Yong, Huang, Jun, Fan, Fan, and Du, You
- Subjects
Computer Science - Computer Vision and Pattern Recognition - Abstract
Non-blind rotary motion deblurring (RMD) aims to recover the latent clear image from a rotary motion blurred (RMB) image. The rotation center is a crucial input parameter in non-blind RMD methods. Existing methods directly estimate the rotation center from the RMB image. However they always suffer significant errors, and the performance of RMD is limited. For the assembled imaging systems, the position of the rotation center remains fixed. Leveraging this prior knowledge, we propose a geometric-based method for rotation center identification and analyze its error range. Furthermore, we construct a RMB imaging system. The experiment demonstrates that our method achieves less than 1-pixel error along a single axis (x-axis or y-axis). We utilize the constructed imaging system to capture real RMB images, and experimental results show that our method can help existing RMD approaches yield better RMD images.
- Published
- 2024
19. Spatio-Temporal Encoding and Decoding-Based Method for Future Human Activity Skeleton Synthesis
- Author
-
Liu, Tingyu, Huang, Jun, and Weng, Chenyi
- Subjects
Computer Science - Computer Vision and Pattern Recognition - Abstract
Inferring future activity information based on observed activity data is a crucial step to improve the accuracy of early activity prediction. Traditional methods based on generative adversarial networks(GAN) or joint learning frameworks can achieve good prediction accuracy under low observation ratios, but they usually have high computational costs. In view of this, this paper proposes a spatio-temporal encoding and decoding-based method for future human activity skeleton synthesis. Firstly, algorithms such as time control, discrete cosine transform, and low-pass filtering are used to cut or pad the skeleton sequences. Secondly, the encoder and decoder are responsible for extracting intermediate semantic encoding from observed skeleton sequences and inferring future sequences from the intermediate semantic encoding, respectively. Finally, joint displacement error, velocity error, and acceleration error, three higher-order kinematic features, are used as key components of the loss function to optimize model parameters. Experimental results show that the proposed future skeleton synthesis algorithm performs better than some existing algorithms. It generates skeleton sequences with smaller errors and fewer model parameters, effectively providing future information for early activity prediction.
- Published
- 2024
20. On the Role of Long-tail Knowledge in Retrieval Augmented Large Language Models
- Author
-
Li, Dongyang, Yan, Junbing, Zhang, Taolin, Wang, Chengyu, He, Xiaofeng, Huang, Longtao, Xue, Hui, and Huang, Jun
- Subjects
Computer Science - Information Retrieval - Abstract
Retrieval augmented generation (RAG) exhibits outstanding performance in promoting the knowledge capabilities of large language models (LLMs) with retrieved documents related to user queries. However, RAG only focuses on improving the response quality of LLMs via enhancing queries indiscriminately with retrieved information, paying little attention to what type of knowledge LLMs really need to answer original queries more accurately. In this paper, we suggest that long-tail knowledge is crucial for RAG as LLMs have already remembered common world knowledge during large-scale pre-training. Based on our observation, we propose a simple but effective long-tail knowledge detection method for LLMs. Specifically, the novel Generative Expected Calibration Error (GECE) metric is derived to measure the ``long-tailness'' of knowledge based on both statistics and semantics. Hence, we retrieve relevant documents and infuse them into the model for patching knowledge loopholes only when the input query relates to long-tail knowledge. Experiments show that, compared to existing RAG pipelines, our method achieves over 4x speedup in average inference time and consistent performance improvement in downstream tasks.
- Published
- 2024
21. Will Southeast Asia be the next global manufacturing hub? A multiway cointegration, causality, and dynamic connectedness analyses on factors influencing offshore decisions
- Author
-
Wang, Haibo, Sua, Lutfu S., Huang, Jun, Ortiz, Jaime, and Alidaee, Bahram
- Subjects
Economics - General Economics ,Statistics - Applications - Abstract
The COVID-19 pandemic has compelled multinational corporations to diversify their global supply chain risk and to relocate their factories to Southeast Asian countries beyond China. Such recent phenomena provide a good opportunity to understand the factors that influenced offshore decisions in the last two decades. We propose a new conceptual framework based on econometric approaches to examine the relationships between these factors. Firstly, the Vector Auto Regression (VAR) for multi-way cointegration analysis by a Johansen test as well as the embedding Granger causality analysis to examine offshore decisions--innovation, technology readiness, infrastructure, foreign direct investment (FDI), and intermediate imports. Secondly, a Quantile Vector Autoregressive (QVAR) model is used to assess the dynamic connectedness among Southeast Asian countries based on the offshore factors. This study explores a system-wide experiment to evaluate the spillover effects of offshore decisions. It reports a comprehensive analysis using time-series data collected from the World Bank. The results of the cointegration, causality, and dynamic connectedness analyses show that a subset of Southeast Asian countries have spillover effects on each other. These countries present a multi-way cointegration and dynamic connectedness relationship. The study contributes to policymaking by providing a data-driven innovative approach through a new conceptual framework., Comment: 30 pages
- Published
- 2024
22. SVASTIN: Sparse Video Adversarial Attack via Spatio-Temporal Invertible Neural Networks
- Author
-
Pan, Yi, Huang, Jun-Jie, Chen, Zihan, Zhao, Wentao, and Wang, Ziyue
- Subjects
Computer Science - Computer Vision and Pattern Recognition - Abstract
Robust and imperceptible adversarial video attack is challenging due to the spatial and temporal characteristics of videos. The existing video adversarial attack methods mainly take a gradient-based approach and generate adversarial videos with noticeable perturbations. In this paper, we propose a novel Sparse Adversarial Video Attack via Spatio-Temporal Invertible Neural Networks (SVASTIN) to generate adversarial videos through spatio-temporal feature space information exchanging. It consists of a Guided Target Video Learning (GTVL) module to balance the perturbation budget and optimization speed and a Spatio-Temporal Invertible Neural Network (STIN) module to perform spatio-temporal feature space information exchanging between a source video and the target feature tensor learned by GTVL module. Extensive experiments on UCF-101 and Kinetics-400 demonstrate that our proposed SVASTIN can generate adversarial examples with higher imperceptibility than the state-of-the-art methods with the higher fooling rate. Code is available at \href{https://github.com/Brittany-Chen/SVASTIN}{https://github.com/Brittany-Chen/SVASTIN}.
- Published
- 2024
23. DAFNet: Dynamic Auxiliary Fusion for Sequential Model Editing in Large Language Models
- Author
-
Zhang, Taolin, Chen, Qizhou, Li, Dongyang, Wang, Chengyu, He, Xiaofeng, Huang, Longtao, Xue, Hui, and Huang, Jun
- Subjects
Computer Science - Computation and Language - Abstract
Recently, while large language models (LLMs) have demonstrated impressive results, they still suffer from hallucination, i.e., the generation of false information. Model editing is the task of fixing factual mistakes in LLMs; yet, most previous works treat it as a one-time task, paying little attention to ever-emerging mistakes generated by LLMs. We address the task of sequential model editing (SME) that aims to rectify mistakes continuously. A Dynamic Auxiliary Fusion Network (DAFNet) is designed to enhance the semantic interaction among the factual knowledge within the entire sequence, preventing catastrophic forgetting during the editing process of multiple knowledge triples. Specifically, (1) for semantic fusion within a relation triple, we aggregate the intra-editing attention flow into auto-regressive self-attention with token-level granularity in LLMs. We further leverage multi-layer diagonal inter-editing attention flow to update the weighted representations of the entire sequence-level granularity. (2) Considering that auxiliary parameters are required to store the knowledge for sequential editing, we construct a new dataset named \textbf{DAFSet}, fulfilling recent, popular, long-tail and robust properties to enhance the generality of sequential editing. Experiments show DAFNet significantly outperforms strong baselines in single-turn and sequential editing. The usage of DAFSet also consistently improves the performance of other auxiliary network-based methods in various scenarios, Comment: ACL2024 findings
- Published
- 2024
24. EasyAnimate: A High-Performance Long Video Generation Method based on Transformer Architecture
- Author
-
Xu, Jiaqi, Zou, Xinyi, Huang, Kunzhe, Chen, Yunkuo, Liu, Bo, Cheng, MengLi, Shi, Xing, and Huang, Jun
- Subjects
Computer Science - Computer Vision and Pattern Recognition ,Computer Science - Computation and Language ,Computer Science - Multimedia - Abstract
This paper presents EasyAnimate, an advanced method for video generation that leverages the power of transformer architecture for high-performance outcomes. We have expanded the DiT framework originally designed for 2D image synthesis to accommodate the complexities of 3D video generation by incorporating a motion module block. It is used to capture temporal dynamics, thereby ensuring the production of consistent frames and seamless motion transitions. The motion module can be adapted to various DiT baseline methods to generate video with different styles. It can also generate videos with different frame rates and resolutions during both training and inference phases, suitable for both images and videos. Moreover, we introduce slice VAE, a novel approach to condense the temporal axis, facilitating the generation of long duration videos. Currently, EasyAnimate exhibits the proficiency to generate videos with 144 frames. We provide a holistic ecosystem for video production based on DiT, encompassing aspects such as data pre-processing, VAE training, DiT models training (both the baseline model and LoRA model), and end-to-end video inference. Code is available at: https://github.com/aigc-apps/EasyAnimate. We are continuously working to enhance the performance of our method., Comment: 8 pages, 6 figures
- Published
- 2024
25. Spin, inclination, and magnetic field evolution of magnetar population in vacuum and plasma-filled magnetospheres
- Author
-
Huang, Jun-Xiang, Lü, Hou-Jun, Rice, Jared, and Liang, En-Wei
- Subjects
Astrophysics - High Energy Astrophysical Phenomena - Abstract
Magnetars are potential energy sources or central engines for numerous transient phenomena in the Universe. How newborn magnetars evolve in different environments remains an open question. Based on both observed and candidate magnetars, it is found that the periods of all magnetars or candidates appear as a bimodal distribution, and are defined as the ``long-P'' and ``short-P'' magnetar subclasses, respectively. We find that for the ``short-P'' subclass of magnetars, the $\dot{P}$ values also appear as a bimodal distribution, and therefore can be classified as ``high-$\dot{P}$ short-P'' and ``low-$\dot{P}$ short-P'' magnetar subclasses. In this paper, we use Monte Carlo simulations to generate synthetic magnetar populations and investigate the evolution of the ``high-$\dot{P}$ short-P'' and ``low-$\dot{P}$ short-P'' magnetar subclasses by considering both the magnetar spin and inclination, as well as the decay of their magnetic field within their evolution in both vacuum and plasma-filled magnetospheres. We find that the magnetar evolution is dependent on both spin and magnetic field, but seems to be insensitive to inclination evolution and magnetospheric environment for the ``high-$\dot{P}$ short-P'' sub-class. In comparison for the case of ``high-$\dot{P}$ short-P'', the magnetar evolution is dependent on spin, magnetic field, and inclination evolution, as well as the magnetospheric environment. The best evolution model should be the case of inclination evolution in vacuum with a small value of $\overline{\mathrm{FOM}}$. The differences in the best-fit parameters also suggest that the ``high-$\dot{P}$ short-P'' and ``low-$\dot{P}$ short-P'' magnetar subclasses may be tracking with different evolution channels., Comment: 18 pages, 2 Tables, and 9 Figures. PRD in press, and matched with the published verison
- Published
- 2024
26. Distilling Instruction-following Abilities of Large Language Models with Task-aware Curriculum Planning
- Author
-
Yue, Yuanhao, Wang, Chengyu, Huang, Jun, and Wang, Peng
- Subjects
Computer Science - Computation and Language - Abstract
The process of instruction tuning aligns pre-trained large language models (LLMs) with open-domain instructions and human-preferred responses. While several studies have explored autonomous approaches to distilling and annotating instructions from more powerful proprietary LLMs, such as ChatGPT, they often neglect the impact of task distributions and the varying difficulty of instructions of the training sets. This oversight can lead to imbalanced knowledge capabilities and poor generalization powers of small student LLMs. To address this challenge, we introduce Task-Aware Curriculum Planning for Instruction Refinement (TAPIR), a multi-round distillation framework with balanced task distributions and dynamic difficulty adjustment. This approach utilizes an oracle LLM to select instructions that are difficult for a student LLM to follow and distill instructions with balanced task distributions. By incorporating curriculum planning, our approach systematically escalates the difficulty levels, progressively enhancing the student LLM's capabilities. We rigorously evaluate TAPIR using two widely recognized benchmarks, including AlpacaEval 2.0 and MT-Bench. The empirical results demonstrate that the student LLMs, trained with our method and less training data, outperform larger instruction-tuned models and strong distillation baselines. The improvement is particularly notable in complex tasks, such as logical reasoning and code generation.
- Published
- 2024
27. Exploring Graph-based Knowledge: Multi-Level Feature Distillation via Channels Relational Graph
- Author
-
Wang, Zhiwei, Huang, Jun, Ma, Longhua, Wu, Chengyu, and Ma, Hongyu
- Subjects
Computer Science - Computer Vision and Pattern Recognition - Abstract
In visual tasks, large teacher models capture essential features and deep information, enhancing performance. However, distilling this information into smaller student models often leads to performance loss due to structural differences and capacity limitations. To tackle this, we propose a distillation framework based on graph knowledge, including a multi-level feature alignment strategy and an attention-guided mechanism to provide a targeted learning trajectory for the student model. We emphasize spectral embedding (SE) as a key technique in our distillation process, which merges the student's feature space with the relational knowledge and structural complexities similar to the teacher network. This method captures the teacher's understanding in a graph-based representation, enabling the student model to more accurately mimic the complex structural dependencies present in the teacher model. Compared to methods that focus only on specific distillation areas, our strategy not only considers key features within the teacher model but also endeavors to capture the relationships and interactions among feature sets, encoding these complex pieces of information into a graph structure to understand and utilize the dynamic relationships among these pieces of information from a global perspective. Experiments show that our method outperforms previous feature distillation methods on the CIFAR-100, MS-COCO, and Pascal VOC datasets, proving its efficiency and applicability.
- Published
- 2024
28. R4: Reinforced Retriever-Reorder-Responder for Retrieval-Augmented Large Language Models
- Author
-
Zhang, Taolin, Li, Dongyang, Chen, Qizhou, Wang, Chengyu, Huang, Longtao, Xue, Hui, He, Xiaofeng, and Huang, Jun
- Subjects
Computer Science - Computation and Language - Abstract
Retrieval-augmented large language models (LLMs) leverage relevant content retrieved by information retrieval systems to generate correct responses, aiming to alleviate the hallucination problem. However, existing retriever-responder methods typically append relevant documents to the prompt of LLMs to perform text generation tasks without considering the interaction of fine-grained structural semantics between the retrieved documents and the LLMs. This issue is particularly important for accurate response generation as LLMs tend to "lose in the middle" when dealing with input prompts augmented with lengthy documents. In this work, we propose a new pipeline named "Reinforced Retriever-Reorder-Responder" (R$^4$) to learn document orderings for retrieval-augmented LLMs, thereby further enhancing their generation abilities while the large numbers of parameters of LLMs remain frozen. The reordering learning process is divided into two steps according to the quality of the generated responses: document order adjustment and document representation enhancement. Specifically, document order adjustment aims to organize retrieved document orderings into beginning, middle, and end positions based on graph attention learning, which maximizes the reinforced reward of response quality. Document representation enhancement further refines the representations of retrieved documents for responses of poor quality via document-level gradient adversarial learning. Extensive experiments demonstrate that our proposed pipeline achieves better factual question-answering performance on knowledge-intensive tasks compared to strong baselines across various public datasets. The source codes and trained models will be released upon paper acceptance., Comment: need to further experiment
- Published
- 2024
29. Guided AbsoluteGrad: Magnitude of Gradients Matters to Explanation's Localization and Saliency
- Author
-
Huang, Jun and Liu, Yan
- Subjects
Computer Science - Computer Vision and Pattern Recognition ,Computer Science - Artificial Intelligence ,Computer Science - Human-Computer Interaction ,Computer Science - Machine Learning - Abstract
This paper proposes a new gradient-based XAI method called Guided AbsoluteGrad for saliency map explanations. We utilize both positive and negative gradient magnitudes and employ gradient variance to distinguish the important areas for noise deduction. We also introduce a novel evaluation metric named ReCover And Predict (RCAP), which considers the Localization and Visual Noise Level objectives of the explanations. We propose two propositions for these two objectives and prove the necessity of evaluating them. We evaluate Guided AbsoluteGrad with seven gradient-based XAI methods using the RCAP metric and other SOTA metrics in three case studies: (1) ImageNet dataset with ResNet50 model; (2) International Skin Imaging Collaboration (ISIC) dataset with EfficientNet model; (3) the Places365 dataset with DenseNet161 model. Our method surpasses other gradient-based approaches, showcasing the quality of enhanced saliency map explanations through gradient magnitude., Comment: CAI2024 Camera-ready Submission
- Published
- 2024
30. Two mixed-ligand coordination polymers based on 2,5-thiophenedicarboxylic acid and flexible N-donor ligands: the protective effect on periodontitis via reducing the release of IL-1β and TNF-α
- Author
-
Wu Shao-Hsuan and Huang Jun-Hui
- Subjects
coordination polymer ,mixed-ligand ,x-ray diffraction ,periodontitis ,Chemistry ,QD1-999 - Abstract
Two novel mixed-ligand coordination polymers, {[Co(tdc)(btrp)]·0.67DMF}n (1) and {[Zn2(bimb)2(tdc)2]·2H2O}n (2) involving 2,5-thiophenedicarboxylate (H2tdc), and bitopic flexible N-donor ligands, 1,3-bis(1,2,4-triazol-1-yl)propane (btrp) and 1,4-bis((1H-benzo[d]imidazol-1-yl)methyl)benzene (bimb), have been synthesized by the hydrothermal method and characterized via IR, elemental analysis, thermal analysis, and powder X-ray diffraction. The biological functional studies were performed; the treatment activity of the compounds on periodontitis and the specific mechanism was explored. First, the real-time RT-PCR was carried out to determine the inflammatory genes nf-κb and p53 relative expression in periodontal mucosal cells after treating with compounds 1 and 2. Then, the level of the inflammatory cytokine in the gingival crevicular fluid after treating with compounds was also determined by the ELISA detection kit.
- Published
- 2020
- Full Text
- View/download PDF
31. Microstructure and mechanical properties of nano-WC reinforced AlSi10Mg fabricated by selective laser melting
- Author
-
YE Han, HUANG Jun-qiang, ZHANG Jian-qiang, LI Cong-cong, and LIU Yong
- Subjects
alsi10mg ,selective laser melting ,nano-wc ,microstructure ,mechanical property ,Materials of engineering and construction. Mechanics of materials ,TA401-492 - Abstract
In order to further enhance the performance of the AlSi10Mg parts fabricated by selective laser melting, the WC/AlSi10Mg composite with 0.1%WC(mass fraction) was obtained by mixing the nano-WC and AlSi10Mg in mixing machine, and the specimen block was fabricated by the selective laser melting machine. By comparing the AlSi10Mg specimens fabricated by the same process, the effects of nano-WC on the microstructure formation and evolution of the microstructure and the mechanical properties of the microstructure were investigated. The results show that the WC/AlSi10Mg powder has good sphericity and the particle size distribution is concentrated in 20-60 μm. The WC/AlSi10Mg sample has the density of over 99% and hardness of about 158.89HV, which is 14.58% higher than that of the AlSi10Mg sample. The WC/AlSi10Mg samples grow uniformly and densely, with obvious molten pool lines. The inside of the crystal grain is α-Al matrix, and the boundary is a eutectic Si phase interposed with WC. The yield strength of the WC/AlSi10Mg sample reaches 337.75 MPa, the ultimate strength is as high as 514.00 MPa, and the elongation is 3.78%. Compared with the same process, AlSi10Mg samples are increased by 4.73%, 6.25% and 35.97%, respectively. Therefore, SLM-processed WC/AlSi10Mg composite parts have better application prospects than AlSi10Mg parts.
- Published
- 2020
32. Effect of LncRNA RP11-173C1.1 Overexpression on Proliferation and Invasion of Thyroid Carcinoma B-CPAP Cells
- Author
-
WANG Qun, YUAN Jie, HUANG Jun, and MEI Hong
- Subjects
thyroid cancer ,rp11-173c1.1 ,proliferation ,invasion ,Neoplasms. Tumors. Oncology. Including cancer and carcinogens ,RC254-282 - Abstract
Objective To detect the expression of lncRNA RP11-173C1.1 in differentiated thyroid tissue, and to investigate the effect of RP11-173C1.1 on the proliferation and invasion of thyroid papillary carcinoma cell line B-CPAP and the molecular mechanism. Methods qRT-PCR was used to detect the expression level of RP11-173C1.1 in differentiated thyroid cancer tissues and thyroid cancer cell lines. The B-CPAP cells with the lowest expression level were used as the research object, and the RP11-173C1.1 overexpression plasmid and the negative control plasmid were transfected into the experimental group and the control group, respectively. The effects of RP11-173C1.1 on the proliferation and invasion of B-CPAP cells were detected by CCK-8 and Transwell invasion assays, respectively. The bioinformatics method predicted the mechanism of RP11-173C1.1. qRT-PCR and Western blot were used to detect the effect of RP11-173C1.1 on downstream gene expression. Results The expression of RP11-173C1.1 was significantly lower in thyroid cancer tissues than that in adjacent tissues (P < 0.01). The expression of RP11-173C1.1 in thyroid cancer cell lines was significantly lower than that in normal thyroid follicular epithelial cells (P < 0.01). Up-regulation of RP11-173C1.1 inhibited the proliferation and invasion abilities of B-CPAP cells (both P < 0.05). RP11-173C1.1 complemented miR-376c-3p, and miR-376c-3p complemented OPCML. Up-regulation of RP11-173C1.1 reduced miR-376c-3p expression and promoted OPCML mRNA and protein expression in B-CPAP cells (all P < 0.01). Conclusion RP11-173C1.1 is lowly expressed in differentiated thyroid carcinoma tissues and cell lines. Up-regulation of RP11-173C1.1 could inhibit the proliferation and invasion of thyroid papillary carcinoma cell line B-CPAP. The possible mechanism is down-regulating miR-376c-3p expression and up-regulating OPCML gene expression.
- Published
- 2020
- Full Text
- View/download PDF
33. Prognostic and Clinicopathological Significance of PD-L1 Expression for Colorectal Cancer: A Meta-analysis
- Author
-
RAO Xionghui, LUO Hongliang, HUANG Jun, and ZHU Zhengming
- Subjects
pd-l1 ,colorectal cancer ,meta-analysis ,prognosis ,Neoplasms. Tumors. Oncology. Including cancer and carcinogens ,RC254-282 - Abstract
Objective To systematically evaluate the prognostic and clinicopathological significance of PD-L1 expression in colorectal cancer(CRC) tissues by meta-analysis. Methods PubMed, Embase, Web of Science, Cochrane Library, CNKI and Wanfang database were used to search for the studies related to PD-L1 expression and the prognosis of CRC. The search time was until June 2018. The data of survival and clinicopathological features were extracted from eligible studies and analyzed by Stata SE12.0 software. Results Eighteen studies with 5724 CRC patients were enrolled. PD-L1 expression in CRC tissues was associated with shorter overall survival (OS) (HR:1.40, 95%CI:1.02-1.93, P=0.039) and relapse-free survival (RFS) (HR:1.67, 95%CI:1.27-2.20, P=0.000), but not with disease-free survival (DFS)(P=0.933). And PD-L1 expression was significantly associated with tumor differentiation(P=0.016) and lymph node metastasis(P=0.028), but not with gender, tumor location, TNM staging, invasion depth, venous invasion, chemotherapy, MSI status or KRAS mutation. In addition, subgroup analysis showed that PD-L1 expression in tumor cells, not in tumor infiltrating immune cells (OS: P=0.991; RFS: P=0.210), was significantly associated with shorter OS(P=0.033) and RFS(P=0.001). Conclusion PD-L1 expression in CRC tissues is associated with poor prognosis. In addition, the prognostic significance of PD-L1 expression in tumor cells might be inconsistent with that in tumor-infiltrating immune cells.
- Published
- 2019
- Full Text
- View/download PDF
34. Mechanical and reactive properties of PTFE/Al/MoO3 composites
- Author
-
HUANG Jun-yi, FANG Xiang, LI Yu-chun, LIU Qiang, WU Shuang-zhang, and SONG Jia-xing
- Subjects
PTFE/Al/MoO3 ,thermite ,quasi-static compression ,dynamic impact ,Materials of engineering and construction. Mechanics of materials ,TA401-492 - Abstract
PTFE/Al/MoO3 composites with different PTFE concentrations were prepared by mold pressing and sintering. The materials were characterized by scanning electron microscopy (SEM) and X-ray diffraction (XRD), in addition, the properties of the composites under quasi-static compression and dynamic impact were also studied. The results indicate that, the strength of PTFE/Al/MoO3 composites is increased first and then decreased with the increase of PTFE, reaches the maximum of 80MPa when the volume fraction of PTFE is 70%. Meanwhile, these specimens can ignite under quasi-static compression conditions, with a huge explosion and bright flames, while no ignition phenomenon was observed in the other specimens under the same conditions. Under the dynamic impact conditions, all of the PTFE/Al/MoO3 composites can ignite, but with the increase of PTFE, the degree of intensity is decreased gradually. The reaction products are AlF3, Al2O3, Mo and C, indicating that the redox reactions of Al and PTFE, Al and MoO3have both been triggered. Under the dual influence of material strength and composition, the energy required for ignition of PTFE/Al/MoO3 composites tends to drop first and then increase with the increase of PTFE.
- Published
- 2019
- Full Text
- View/download PDF
35. Relationship between the Number of Neutrophils and Myofibroblasts during Diabetic Wound Healing and Wound Age
- Author
-
HUANG Jun-jie, YAO Yi, XIA Chong-Jian,et al.
- Subjects
forensic pathology ,diabetes mellitus ,wound healing ,neutrophil ,myofibroblast ,wound age estimation ,mice ,Medicine - Abstract
Objective To investigate the sequential changes of the number of neutrophils and myofibroblasts during diabetic wound healing, and discuss its application value in wound age estimation. Methods Diabetic DB mice and mice of the same age in the normal control group were selected, a wound healing model was established, wound samples were taken at different time points, while the number of neutrophils and myofibroblasts during diabetic wound healing were determined by immunohistochemical staining technique. Results The number of infiltrated neutrophils in the wounds of control and diabetic groups reached the peak respectively at 12 h and 5 d after injury. Compared with the control group, the number of neutrophils in the diabetic group decreased significantly from 6 h to 1 d after injury, but increased markedly from 5 d to 14 d. From 5 d to 10 d after injury, the average number of neutrophils at high magnification in wounds of the diabetic group was over 30, while that of neutrophils in wounds of the control group was less than 20. Myofibroblasts appeared in wounds from 3 d to 14 d after injury in the control group and from 5 d to 14 d after injury in the diabetic group. The difference in the number of myofibroblasts in wounds between control group and diabetic group from 3 to 7 d after injury had statistical significance. Conclusion In comparison with normal wound healing, the number of neutrophils and myofibroblasts during diabetic wound healing shows different sequential changes. The results of this study can provide reference for wound age estimation of patients with severe diabetes.
- Published
- 2019
- Full Text
- View/download PDF
36. Development and evaluation of a differential diagnostic test strip for antibodies induced by HIV-1 vaccine and natural infections
- Author
-
YUAN Zuowei, WANG Yonghong, HUANG Jun, and YE Yangmin
- Subjects
hiv-1 ,natural infection ,vaccine ,antibody ,differential diagnosis ,diagnostic reagent kits ,Medicine (General) ,R5-920 - Abstract
Objective To develop a differential diagnostic kit for distinguishing antibodies induced by HIV-1 infection and HIV-1 vaccine and evaluate its diagnostic performance. Methods Nitrocellulose membrane was coated with HIV-1 envelope protein gp41 and HIV-1 recognition fusion peptide SK1-SK2-SK3 obtained using a prokaryotic expression system. The differential diagnostic test strip for HIV-1 antibodies was prepared by assembling the coated membrane with a non-woven fabric coated with protein A-coupled colloidal gold. The performance of the test strip was evaluated using samples of HIV-positive sera, HIV-negative sera, HIV serum panel, HIV Tiantan vaccine sera and other cross-reactive sera, and the test results using this strip were compared with those by ELISA. Results The sensitivity of SK1-SK2-SK3 coating on the novel differential diagnostic test strip for HIV-1 antibody detection was 93.5%, with a detection specificity of 100.0%. When combined with gp41 coating, the test strip showed a diagnostic sensitivity of 100.0%. The test results with the strip were consistent with those by ELISA (P>0.05). For the Tiantan vaccine serum, the SK1-SK2-SK3 coating on the strip all yielded negative results without the false positive results of the gp41 coating. Conclusion This novel HIV-1 antibody diagnostic test strip can effectively distinguish the antibodies produced by HIV-1 infection from those generated by HIV-1 vaccination without false positive results as are common with the currently available commercial HIV test kits.
- Published
- 2019
- Full Text
- View/download PDF
37. Microstructure and mechanical properties of nano-WC reinforced AlSi10Mg fabricated by selective laser melting
- Author
-
YE Han, HUANG Jun-qiang, ZHANG Jian-qiang, LI Cong-cong, and LIU Yong
- Subjects
alsi10mg ,selective laser melting ,nano-wc ,microstructure ,mechanical property ,Materials of engineering and construction. Mechanics of materials ,TA401-492 - Abstract
In order to further enhance the performance of the AlSi10Mg parts fabricated by selective laser melting, the WC/AlSi10Mg composite with 0.1%WC(mass fraction) was obtained by mixing the nano-WC and AlSi10Mg in mixing machine, and the specimen block was fabricated by the selective laser melting machine. By comparing the AlSi10Mg specimens fabricated by the same process, the effects of nano-WC on the microstructure formation and evolution of the microstructure and the mechanical properties of the microstructure were investigated. The results show that the WC/AlSi10Mg powder has good sphericity and the particle size distribution is concentrated in 20-60 μm. The WC/AlSi10Mg sample has the density of over 99% and hardness of about 158.89HV, which is 14.58% higher than that of the AlSi10Mg sample. The WC/AlSi10Mg samples grow uniformly and densely, with obvious molten pool lines. The inside of the crystal grain is α-Al matrix, and the boundary is a eutectic Si phase interposed with WC. The yield strength of the WC/AlSi10Mg sample reaches 337.75 MPa, the ultimate strength is as high as 514.00 MPa, and the elongation is 3.78%. Compared with the same process, AlSi10Mg samples are increased by 4.73%, 6.25% and 35.97%, respectively. Therefore, SLM-processed WC/AlSi10Mg composite parts have better application prospects than AlSi10Mg parts.
- Published
- 2019
38. A strategy for human-robot collaboration in taking products apart for remanufacture
- Author
-
Huang Jun, Pham Duc Truong, Wang Yongjing, Ji Chunqian, Xu Wenjun, Liu Quan, and Zhou Zude
- Subjects
industry 4.0 ,remanufacturing ,robotic disassembly ,human-robot collaboration ,collaborative robot ,active compliance control ,Engineering (General). Civil engineering (General) ,TA1-2040 ,Mechanics of engineering. Applied mechanics ,TA349-359 - Abstract
Using humans and robots working together to disassemble end-of-life (EoL) products so that they can be remanufactured enables catering for the effects of uncertainties in the condition of those products and unpredictability in remanufacturing operations. Collaborative robots are one of the key elements of Industry 4.0. This paper presents a strategy for human-robot collaborative disassembly that exploits the active compliance control capability of collaborative robots. The paper outlines the proposed human-robot collaboration strategy and active compliance control. It then presents a case study of robotic disassembly designed to show a practical implementation of the proposed strategy to take products apart. The study involved the dismantling of an automotive water pump by a collaborative industrial robot working with a human operator and by two collaborative robots and a human operator. Active compliance control was implemented to separate press-fitted components from the water pump. The results demonstrate the feasibility of the strategy and its effectiveness at increasing flexibility and adaptability by reducing setup efforts.
- Published
- 2019
39. A revamped rat reference genome improves the discovery of genetic diversity in laboratory rats
- Author
-
de Jong, Tristan V, Pan, Yanchao, Rastas, Pasi, Munro, Daniel, Tutaj, Monika, Akil, Huda, Benner, Chris, Chen, Denghui, Chitre, Apurva S, Chow, William, Colonna, Vincenza, Dalgard, Clifton L, Demos, Wendy M, Doris, Peter A, Garrison, Erik, Geurts, Aron M, Gunturkun, Hakan M, Guryev, Victor, Hourlier, Thibaut, Howe, Kerstin, Huang, Jun, Kalbfleisch, Ted, Kim, Panjun, Li, Ling, Mahaffey, Spencer, Martin, Fergal J, Mohammadi, Pejman, Ozel, Ayse Bilge, Polesskaya, Oksana, Pravenec, Michal, Prins, Pjotr, Sebat, Jonathan, Smith, Jennifer R, Woods, Leah C Solberg, Tabakoff, Boris, Tracey, Alan, Uliano-Silva, Marcela, Villani, Flavia, Wang, Hongyang, Sharp, Burt M, Telese, Francesca, Jiang, Zhihua, Saba, Laura, Wang, Xusheng, Murphy, Terence D, Palmer, Abraham A, Kwitek, Anne E, Dwinell, Melinda R, Williams, Robert W, Li, Jun Z, and Chen, Hao
- Subjects
Biological Sciences ,Bioinformatics and Computational Biology ,Genetics ,Human Genome ,Biotechnology ,Good Health and Well Being ,Rats ,Animals ,Genome ,Molecular Sequence Annotation ,Genomics ,Whole Genome Sequencing ,Genetic Variation ,Rnor_6.0 ,genetic map ,heterogeneous stock ,hybrid rat diversity panel ,inbred strains ,mRatBN7.2 ,phylogenetic tree ,rat ,recombinant inbred ,reference genome - Abstract
The seventh iteration of the reference genome assembly for Rattus norvegicus-mRatBN7.2-corrects numerous misplaced segments and reduces base-level errors by approximately 9-fold and increases contiguity by 290-fold compared with its predecessor. Gene annotations are now more complete, improving the mapping precision of genomic, transcriptomic, and proteomics datasets. We jointly analyzed 163 short-read whole-genome sequencing datasets representing 120 laboratory rat strains and substrains using mRatBN7.2. We defined ∼20.0 million sequence variations, of which 18,700 are predicted to potentially impact the function of 6,677 genes. We also generated a new rat genetic map from 1,893 heterogeneous stock rats and annotated transcription start sites and alternative polyadenylation sites. The mRatBN7.2 assembly, along with the extensive analysis of genomic variations among rat strains, enhances our understanding of the rat genome, providing researchers with an expanded resource for studies involving rats.
- Published
- 2024
40. AlphaFin: Benchmarking Financial Analysis with Retrieval-Augmented Stock-Chain Framework
- Author
-
Li, Xiang, Li, Zhenyu, Shi, Chen, Xu, Yong, Du, Qing, Tan, Mingkui, Huang, Jun, and Lin, Wei
- Subjects
Computer Science - Computation and Language - Abstract
The task of financial analysis primarily encompasses two key areas: stock trend prediction and the corresponding financial question answering. Currently, machine learning and deep learning algorithms (ML&DL) have been widely applied for stock trend predictions, leading to significant progress. However, these methods fail to provide reasons for predictions, lacking interpretability and reasoning processes. Also, they can not integrate textual information such as financial news or reports. Meanwhile, large language models (LLMs) have remarkable textual understanding and generation ability. But due to the scarcity of financial training datasets and limited integration with real-time knowledge, LLMs still suffer from hallucinations and are unable to keep up with the latest information. To tackle these challenges, we first release AlphaFin datasets, combining traditional research datasets, real-time financial data, and handwritten chain-of-thought (CoT) data. It has a positive impact on training LLMs for completing financial analysis. We then use AlphaFin datasets to benchmark a state-of-the-art method, called Stock-Chain, for effectively tackling the financial analysis task, which integrates retrieval-augmented generation (RAG) techniques. Extensive experiments are conducted to demonstrate the effectiveness of our framework on financial analysis., Comment: COLING 2024. The first three authors contributed equally. Project website: https://github.com/AlphaFin-proj/AlphaFin
- Published
- 2024
41. TRELM: Towards Robust and Efficient Pre-training for Knowledge-Enhanced Language Models
- Author
-
Yan, Junbing, Wang, Chengyu, Zhang, Taolin, He, Xiaofeng, Huang, Jun, Huang, Longtao, Xue, Hui, and Zhang, Wei
- Subjects
Computer Science - Computation and Language - Abstract
KEPLMs are pre-trained models that utilize external knowledge to enhance language understanding. Previous language models facilitated knowledge acquisition by incorporating knowledge-related pre-training tasks learned from relation triples in knowledge graphs. However, these models do not prioritize learning embeddings for entity-related tokens. Moreover, updating the entire set of parameters in KEPLMs is computationally demanding. This paper introduces TRELM, a Robust and Efficient Pre-training framework for Knowledge-Enhanced Language Models. We observe that entities in text corpora usually follow the long-tail distribution, where the representations of some entities are suboptimally optimized and hinder the pre-training process for KEPLMs. To tackle this, we employ a robust approach to inject knowledge triples and employ a knowledge-augmented memory bank to capture valuable information. Furthermore, updating a small subset of neurons in the feed-forward networks (FFNs) that store factual knowledge is both sufficient and efficient. Specifically, we utilize dynamic knowledge routing to identify knowledge paths in FFNs and selectively update parameters during pre-training. Experimental results show that TRELM reduces pre-training time by at least 50% and outperforms other KEPLMs in knowledge probing tasks and multiple knowledge-aware language understanding tasks.
- Published
- 2024
42. DiffChat: Learning to Chat with Text-to-Image Synthesis Models for Interactive Image Creation
- Author
-
Wang, Jiapeng, Wang, Chengyu, Cao, Tingfeng, Huang, Jun, and Jin, Lianwen
- Subjects
Computer Science - Computation and Language ,Computer Science - Computer Vision and Pattern Recognition - Abstract
We present DiffChat, a novel method to align Large Language Models (LLMs) to "chat" with prompt-as-input Text-to-Image Synthesis (TIS) models (e.g., Stable Diffusion) for interactive image creation. Given a raw prompt/image and a user-specified instruction, DiffChat can effectively make appropriate modifications and generate the target prompt, which can be leveraged to create the target image of high quality. To achieve this, we first collect an instruction-following prompt engineering dataset named InstructPE for the supervised training of DiffChat. Next, we propose a reinforcement learning framework with the feedback of three core criteria for image creation, i.e., aesthetics, user preference, and content integrity. It involves an action-space dynamic modification technique to obtain more relevant positive samples and harder negative samples during the off-policy sampling. Content integrity is also introduced into the value estimation function for further improvement of produced images. Our method can exhibit superior performance than baseline models and strong competitors based on both automatic and human evaluations, which fully demonstrates its effectiveness.
- Published
- 2024
43. Improved theoretical prediction of nanoparticle sizes with the resistive-pulse technique
- Author
-
Gao, Zihao, Ma, Long, Liu, Zhe, Huang, Jun, Liu, Hanlian, Huang, Chuanzhen, and Qiu, Yinghua
- Subjects
Physics - Chemical Physics - Abstract
With the resistive-pulse technique (RPT), nanopores serve as the nanofluidic sensors of various analytes for their many physical and chemical properties. Here, we focus on the size measurement and its theoretical prediction for sub-200 nm nanoparticles with RPT. Through systematical investigation of the current blockade of nanoparticles across cylindrical nanopores with simulations, Maxwell method considering the shape coefficient and access resistances agrees well with simulation results. However, the widely used integration method of the resistance has distinct deviations in various cases. With the introduction of a correction factor \b{eta} to the integration method, our revised equations can provide good predictions for simulation results. \b{eta} shows a strong dependence on the diameter ratio (d over D) of the nanoparticle and nanopore. Following the same strategy, modified equations are provided for the accurate size prediction for nanoparticles across conical nanopores, where the integration method is the default convenient way. The correction factor \b{eta}' relates to \b{eta} in cylindrical nanopores. \b{eta}' exhibits independence on the pore geometry parameters and diameters of nanoparticles, but dependence on the surface charge density of conical nanopores. Our improved equations can provide theoretical predictions for the accurate size detection of 100-200 nm diameter nanoparticles across cylindrical and conical nanopores., Comment: 29 pages 7 figures
- Published
- 2024
- Full Text
- View/download PDF
44. Towards Understanding Cross and Self-Attention in Stable Diffusion for Text-Guided Image Editing
- Author
-
Liu, Bingyan, Wang, Chengyu, Cao, Tingfeng, Jia, Kui, and Huang, Jun
- Subjects
Computer Science - Computer Vision and Pattern Recognition - Abstract
Deep Text-to-Image Synthesis (TIS) models such as Stable Diffusion have recently gained significant popularity for creative Text-to-image generation. Yet, for domain-specific scenarios, tuning-free Text-guided Image Editing (TIE) is of greater importance for application developers, which modify objects or object properties in images by manipulating feature components in attention layers during the generation process. However, little is known about what semantic meanings these attention layers have learned and which parts of the attention maps contribute to the success of image editing. In this paper, we conduct an in-depth probing analysis and demonstrate that cross-attention maps in Stable Diffusion often contain object attribution information that can result in editing failures. In contrast, self-attention maps play a crucial role in preserving the geometric and shape details of the source image during the transformation to the target image. Our analysis offers valuable insights into understanding cross and self-attention maps in diffusion models. Moreover, based on our findings, we simplify popular image editing methods and propose a more straightforward yet more stable and efficient tuning-free procedure that only modifies self-attention maps of the specified attention layers during the denoising process. Experimental results show that our simplified method consistently surpasses the performance of popular approaches on multiple datasets.
- Published
- 2024
45. The classification of two-distance transitive dihedrants
- Author
-
Huang, Jun-Jie, Feng, Yan-Quan, Zhou, Jin-Xin, and Yin, Fu-Gang
- Subjects
Mathematics - Combinatorics ,05C25, 20B15, 20B30 - Abstract
A vertex transitive graph $\Gamma$ is said to be $2$-distance transitive if for each vertex $u$, the group of automorphisms of $\Gamma$ fixing the vertex $u$ acts transitively on the set of vertices at distance $1$ and $2$ from $u$, while $\Gamma$ is said to be $2$-arc transitive if its automorphism group is transitive on the set of $2$-arcs. Then $2$-arc transitive graphs are $2$-distance transitive. The classification of $2$-arc transitive Cayley graphs on dihedral groups was given by Du, Malni\v{c} and Maru\v{s}i\v{c} in [Classification of 2-arc-transitive dihedrants, J. Combin. Theory Ser. B 98 (2008), 1349--1372]. In this paper, it is shown that a connected 2-distance transitive Cayley graph on the dihedral group of order $2n$ is either $2$-arc transitive, or isomorphic to the complete multipartite graph $K_{m[b]}$ for some $m\geq3$ and $b\geq2$ with $mb=2n$., Comment: 21
- Published
- 2024
46. Exploring Advanced Methodologies in Security Evaluation for LLMs
- Author
-
Huang, Jun, Zhang, Jiawei, Wang, Qi, Han, Weihong, and Zhang, Yanchun
- Subjects
Computer Science - Cryptography and Security - Abstract
Large Language Models (LLMs) represent an advanced evolution of earlier, simpler language models. They boast enhanced abilities to handle complex language patterns and generate coherent text, images, audios, and videos. Furthermore, they can be fine-tuned for specific tasks. This versatility has led to the proliferation and extensive use of numerous commercialized large models. However, the rapid expansion of LLMs has raised security and ethical concerns within the academic community. This emphasizes the need for ongoing research into security evaluation during their development and deployment. Over the past few years, a substantial body of research has been dedicated to the security evaluation of large-scale models. This article an in-depth review of the most recent advancements in this field, providing a comprehensive analysis of commonly used evaluation metrics, advanced evaluation frameworks, and the routine evaluation processes for LLMs. Furthermore, we also discuss the future directions for advancing the security evaluation of LLMs.
- Published
- 2024
47. Do Large Language Models Understand Logic or Just Mimick Context?
- Author
-
Yan, Junbing, Wang, Chengyu, Huang, Jun, and Zhang, Wei
- Subjects
Computer Science - Computation and Language ,Computer Science - Artificial Intelligence - Abstract
Over the past few years, the abilities of large language models (LLMs) have received extensive attention, which have performed exceptionally well in complicated scenarios such as logical reasoning and symbolic inference. A significant factor contributing to this progress is the benefit of in-context learning and few-shot prompting. However, the reasons behind the success of such models using contextual reasoning have not been fully explored. Do LLMs have understand logical rules to draw inferences, or do they ``guess'' the answers by learning a type of probabilistic mapping through context? This paper investigates the reasoning capabilities of LLMs on two logical reasoning datasets by using counterfactual methods to replace context text and modify logical concepts. Based on our analysis, it is found that LLMs do not truly understand logical rules; rather, in-context learning has simply enhanced the likelihood of these models arriving at the correct answers. If one alters certain words in the context text or changes the concepts of logical terms, the outputs of LLMs can be significantly disrupted, leading to counter-intuitive responses. This work provides critical insights into the limitations of LLMs, underscoring the need for more robust mechanisms to ensure reliable logical reasoning in LLMs.
- Published
- 2024
48. Sliding Mode Observers for Set-valued Lur'e Systems with Uncertainties Beyond Observational Range
- Author
-
Adly, Samir, Huang, Jun, and Le, Ba Khiet
- Subjects
Mathematics - Optimization and Control - Abstract
In this paper, we introduce a new sliding mode observer for Lur'e set-valued dynamical systems, particularly addressing challenges posed by uncertainties not within the standard range of observation. Traditionally, most of Luenberger-like observers and sliding mode observer have been designed only for uncertainties in the range of observation. Central to our approach is the treatment of the uncertainty term which we decompose into two components: the first part in the observation subspace and the second part in its complemented subspace. We establish that when the second part converges to zero, an exact sliding mode observer for the system can be obtained. In scenarios where this convergence does not occur, our methodology allows for the estimation of errors between the actual state and the observer state. This leads to a practical interval estimation technique, valuable in situations where part of the uncertainty lies outside the observable range. Finally, we show that our observer is also a T- observer as well as a strong H-infinity observer.
- Published
- 2024
49. Diffutoon: High-Resolution Editable Toon Shading via Diffusion Models
- Author
-
Duan, Zhongjie, Wang, Chengyu, Chen, Cen, Qian, Weining, and Huang, Jun
- Subjects
Computer Science - Computer Vision and Pattern Recognition - Abstract
Toon shading is a type of non-photorealistic rendering task of animation. Its primary purpose is to render objects with a flat and stylized appearance. As diffusion models have ascended to the forefront of image synthesis methodologies, this paper delves into an innovative form of toon shading based on diffusion models, aiming to directly render photorealistic videos into anime styles. In video stylization, extant methods encounter persistent challenges, notably in maintaining consistency and achieving high visual quality. In this paper, we model the toon shading problem as four subproblems: stylization, consistency enhancement, structure guidance, and colorization. To address the challenges in video stylization, we propose an effective toon shading approach called \textit{Diffutoon}. Diffutoon is capable of rendering remarkably detailed, high-resolution, and extended-duration videos in anime style. It can also edit the content according to prompts via an additional branch. The efficacy of Diffutoon is evaluated through quantitive metrics and human evaluation. Notably, Diffutoon surpasses both open-source and closed-source baseline approaches in our experiments. Our work is accompanied by the release of both the source code and example videos on Github (Project page: https://ecnu-cilab.github.io/DiffutoonProjectPage/).
- Published
- 2024
50. From Perspective of Hippocampal Plasticity: Function of Antidepressant Chinese Medicine Xiaoyaosan
- Author
-
Zhang, Wu-jing, Guo, Ze-xuan, Wang, Yi-di, Fang, Shao-yi, Wan, Chun-miao, Yu, Xiao-long, Guo, Xiao-fang, Chen, Yue-yue, Zhou, Xuan, Huang, Jun-qing, Li, Xiao-juan, Chen, Jia-xu, and Fan, Li-li
- Published
- 2024
- Full Text
- View/download PDF
Catalog
Discovery Service for Jio Institute Digital Library
For full access to our library's resources, please sign in.