6,239 results on '"Wang Hongyu"'
Search Results
2. Bogie Bracket Random Vibration Simulation Analysis and Structural Improvement Based on Field-measured Spectrum
- Author
-
XIE Yingying, WANG Hongyu, and JIN Xin
- Subjects
urban rail transit ,random vibration ,field-measured spectrum ,fatigue damage ,structural improvement ,Transportation engineering ,TA1001-1280 - Abstract
Objective It is aimed to investigate the fatigue damage of an antenna bracket structure installed at urban rail transit vehicle bogie frame end under random vibration excitation, and improvement measures are proposed. Method Finite element modeling of the antenna bracket system is conducted using finite element analysis software. Modal analysis of the structure is performed, and the on-site measured vibration ASD (acceleration spectral density) from the track is used as the load excitation instead of the standard spectrum. A random vibration simulation of the antenna bracket is conducted using fatigue analysis software nCode to obtain the bracket fatigue damage result. Improvement measures are proposed for the fatigue weak points of the bracket, and the effectiveness of the improvement measures is verified. Result & Conclusion Simulation results reveal that the fatigue weak points of this structure are at the roots of the reinforcement plates on bracket both sides, with a maximum total damage of 15.58, failing to meet application requirements. The improvement measures involve locally thickened plate roots on bracket both sides and adding support ribs on the sides. After implementing these measures, the fatigue damage value of the bracket significantly decreased, with a maximum total damage of 0.02, meeting application requirements, thereby verifying the effectiveness of the improvement scheme.
- Published
- 2024
- Full Text
- View/download PDF
3. Molecular mechanism and intervention measures of microvascular complications in diabetes
- Author
-
Xu Rui, Fang Ziming, Wang Hongyu, Gu Ye, Yu Liying, Zhang Boyang, and Xu Jingyu
- Subjects
microvascular complication ,epidemiology ,molecular mechanism ,risk factors ,treatment ,Medicine - Abstract
In this article, the epidemiology, molecular mechanism of occurrence and development, risk factors, and treatment of diabetic microvascular complications such as diabetic nephropathy, diabetic retinopathy, and diabetic peripheral neuropathy were discussed, providing the theoretical basis for more accurate elucidation of the pathogenesis and treatment of diabetic microvascular complications.
- Published
- 2024
- Full Text
- View/download PDF
4. Research and application progress of high entropy alloy materials in laser additive repair
- Author
-
LI Zhanming, WANG Menglu, SUN Xiaofeng, WANG Rui, and WANG Hongyu
- Subjects
high entropy alloy ,composite design ,process optimization ,strengthening mechanism ,laser additive repair ,Motor vehicles. Aeronautics. Astronautics ,TL1-4050 - Abstract
Introducing emerging high entropy alloys materials into advanced intelligent manufacturing for laser additive repair is expected to promote the deep integration of new generation of materials and manufacturing technology and greatly improve the utilization of raw materials and energy,which have broad application fields and excellent development prospects. This paper introduces the application status of high entropy alloys in laser additive repair,and points out that the mismatch of strength and toughness,inaccurate performance control and unclear strengthening mechanism are the key scientific problems that need to be solved urgently in the expansion and application of high entropy alloys in laser additive repair. Exploring the ductile-brittle transition mechanism of the high entropy alloys cladding coating metal,clarifying the basic mapping relationship among the materials,processes,microstructure and coating performance of cladding coatings,obtaining a complete and effective method for predicting the composition of high entropy alloys,innovating the design of alloy powder system,optimizing and adjusting the control processes,and obtaining a high-performance cladding coating suitable for extreme service environment and with low cost are the main research focus and development trends in the future.
- Published
- 2024
- Full Text
- View/download PDF
5. Experimental and simulation research on hollow AZ31 magnesium alloy three-channel joint by hot extrusion forming with sand mandrel
- Author
-
Shi Shengnan, Wang Hongyu, Teng Fei, Jiang Lei, Sun Juncai, Sun Jie, and Zhang Shunhu
- Subjects
AZ31 magnesium alloy ,Three-channel joint ,Sand ,Experiments and the finite element ,Die angle ,Mining engineering. Metallurgy ,TN1-997 - Abstract
Magnesium alloy is one of the lightest metal structural materials. The weight is further reduced through the hollow structure. However, the hollow structure is easily damaged during processing. In order to maintain the hollow structure and to transfer the stresses during the high temperature deformation, the sand mandrel is proposed. In this paper, the hollow AZ31 magnesium alloy three-channel joint is studied by hot extrusion forming. Sand as one of solid granule medium is used to fill the hollow magnesium alloy. The extrusion temperatures are 230 °C and 300 °C, respectively. The process parameters (die angle, temperature, bottom thickness, sidewall thickness, edge-to-middle ratio in bottom, bottom shape) of the hollow magnesium alloy are analyzed based on the results of experiments and the finite element method. The results are shown that the formability of the hollow magnesium alloy will be much better when the ratio of sidewall thickness to the bottom thickness is 1:1.5. Also when edge-to-middle ratio in bottom is about 1:1.5, a better forming product can be received. The best bottom shape in these experiments will be convex based on the forming results. The grain will be refined obviously after the extrusion. Also the microstructures will be shown as streamlines. And these lines will be well agreement with the mold in the corner.
- Published
- 2024
- Full Text
- View/download PDF
6. Simulation of internal electric field of particle source water absorbed dose absolute measurement device
- Author
-
HANG Zhongbin, LIU Yuntao, SONG Mingzhe, WEI Kexin, WANG Hongyu, LIU Chuanfeng, TENG Zhongbin, and GENG Xuan
- Subjects
electric field simulation ,finite element analysis ,particle source ,extrapolation ionization chamber ,Nuclear engineering. Atomic power ,TK9001-9401 - Abstract
BackgroundThe physical quantity concerned in the clinical application of 125I particle source brachytherapy is the 1 cm water absorbed dose rate D˙w,1 cm. However, there is no corresponding standard for this physical quantity in China. It is planned to develop an ionization chamber for the absolute measurement of the water absorbed dose rate of the 125I particle source as a standard device, hence the internal electric field of the ionization chamber must first be analyzed to obtain the most reasonable design scheme to satisfy the condition that the electric field intensity is uniformly distributed in the ionization chamber.PurposeThis study aims to simulate the internal electric field of the ionization chamber for the design of particle source water absorbed dose absolute measurement device.MethodsFirstly, the internal model of the ionization chamber was established by Maxwell software to simulate the distribution of electric field intensity under the six variables of the ionization chamber: with or without a protective electrode, different protective electrode ring width, different insulation ring width, different grid number, different grid shape and different grid thickness. Then, qualitative and quantitative analysis were carried out using finite element method. Finally, the influence of different variables on the distribution of electric field intensity in the ionization chamber of the absolute measurement of water absorbed dose of 125I particle source was obtained.ResultsAnalysis results show the ratio of the width of the guard electrode to the radius of the collector electrode must be not less than 2 for design of the ionization chamber. The edge effect at the edge of the collector is increased with width increment of the insulator ring, hence the width of the insulator ring should be reduced as much as possible. When the number of grid electrodes is 15, the variation of the electric field intensity can be reduced to about 1%. When the grid section is rectangular, the variation in electric field intensity is small compared with the circular and triangular grid sections. The larger the thickness of the grid, the more severe the edge effect at the edge of the grid, and the thickness of the grid should be reduced as much as possible.ConclusionsThe uniformity of the electric field can be effectively improved by increasing the number of grid electrodes. Results of this study is helpful to optimal design of the standard device for measuring the absolute water absorbed dose of 125I particle source.
- Published
- 2024
- Full Text
- View/download PDF
7. Color Processing System of Art Color Based on Fractional Differential Equation
- Author
-
Wang Hongyu, Cheng Long, and Huang Zheng
- Subjects
color repair ,fractional differential equation ,image processing technology ,bridging edge repair ,color extraction ,fix code settings ,34a08 ,Mathematics ,QA1-939 - Abstract
This paper mainly studies a high-precision correction method for warp knitting seamless images. The color factor conversion method is used to transform the gray level of the center point in the frequency domain. Then this paper proposes an image-denoising method based on an average way to improve image quality. The multichannel scalar channel is solved using the gradient domain method, and the fractional differential equations of two images are given. A structure tensor matrix is constructed to describe the geometric properties of images. The regional contrast of the image is calculated by using the characteristics of adjacent layers of pixels. The fusion of multiple images is based on the limitations of object functionality. This can make the texture structure of artistic color pictures more obvious. The experiment of color restoration shows that this method has a good interpretation effect. The color image processing technique has obtained an excellent restoration effect.
- Published
- 2023
- Full Text
- View/download PDF
8. PMU abnormal data identification algorithm based on stream clustering
- Author
-
DENG Xiaoyu, WANG Xiangbing, CAO Huazhen, WANG Liuhuo, YAN Hongfeng, and WANG Hongyu
- Subjects
phasor measurement unit (pmu) ,abnormal data ,event data ,identification framework ,information entropy ,stream clustering ,Applications of electric power ,TK4001-4102 - Abstract
In order to ensure the accurate application of the data collected by the phasor measurement unit (PMU), it is necessary to eliminate the abnormal data in its measured values. The existing PMU abnormal data identification algorithm has the disadvantages of high algorithm complexity, difficulty in online updating, difficulty in the calibration of multi-source data, and difficulty in application relying on multi-source data. In this paper, an abnormal data identification framework is proposed based on the PMU event data and abnormal data model and the definition of PMU abnormal data identification information entropy. On the basis of the framework, a PMU abnormal data identification algorithm is proposed based on the balanced iterative reducing and clustering using hierarchies (BIRCH) algorithm. The proposed algorithm is implemented, and an algorithm experiment is carried out for the PMU dataset of a substation. The experimental results show that the proposed algorithm has better accuracy and real-time performance than one-class support vector machine (OCSVM) algorithm and gap statistic algorithm.
- Published
- 2023
- Full Text
- View/download PDF
9. In-situ zircon U-Pb dating method by LA-ICP-MS and discussions on the effect of different beam spot diameters on the dating results
- Author
-
WANG Sen, ZHANG Shuanhong, ZHANG Qiqi, LIANG Xia, KONG Linghao, HU Guohui, WANG Kai, WANG Hongyu, and CAI Yuhang
- Subjects
la-icp-ms ,zircon ,u-pb dating ,trace element ,Geology ,QE1-996.5 - Abstract
By using the newly introduced GeoLas HD 193 nm ArF excimer laser ablation system and the Agilent 7900 quadrupole inductively coupled plasma mass spectrometer (ICP-MS) in the Key Laboratory of Paleomagnetism and Tectonic Reconstruction at the Institute of Geomechanics, we successfully established the in-situ zircon U-Pb dating method and trace element analytical method by LA-ICP-MS. U-Pb dating experiments were carried out on Plešovice, Temora1 and Qinghu zircons under the experimental conditions of 32 μm beam spot diameter, 5.0 J/cm2 energy density and 5 Hz denudation frequency using the standard zircon 91500 as the external standard. The ideal age results are consistent with their recommended values, and the measured U-Pb ages for Plešovice zircon remain stable in different periods. Meanwhile, a comparative study of the U-Pb dating results in different laboratories was carried out on an unknown sample 11-5, and the dating results are consistent with that obtained in the State Key Laboratory of Geological Processes and Mineral Resources at China University of Geosciences (Wuhan) within analytical error. Trace elements in zircon standard 91500 and glass standard NIST SRM 612 were analyzed by using NIST SRM 610 as an external standard and 29Si as an internal standard, and the measured values are consistent with the recommended values of trace element contents. In addition, the influence of different beam spot diameters on the zircon U-Pb age results was explored by using the Plešovice zircon as an example. The results show that a denudation diameter of 16~44 μm can obtain reliable zircon U-Pb ages similar to the recommended age, and under the same energy density and denudation frequency conditions, U-Pb ages obtained by using 32~44 μm beam spot diameters are more accurate than those obtained by 16~24 μm beam spot diameters.
- Published
- 2022
- Full Text
- View/download PDF
10. Design and preparation of an equivalent radioisotope heat source and its performance test
- Author
-
WANG Hongyu, XU Zhiheng, WANG Jiyu, LIU Shichao, LU Hongbo, LI Xinyi, LIU Yunpeng, and TANG Xiaobin
- Subjects
radioisotope ,simulated heat source ,comsol simulation ,general-purpose heat source ,radioisotope thermoelectric generator ,Nuclear engineering. Atomic power ,TK9001-9401 - Abstract
BackgroundA general-purpose heat source (GPHS) is the most established heat source module for radioisotope thermoelectric generators (RTGs) and a key reference for the equivalence of electrically heated analog heat sources during the development and testing of radioisotope power supplies.PurposeThe study aims to develop a highly efficient electric heating simulation heat source to meet the requirements for equivalent testing and verification of non-nuclear units of radioisotope power systems.MethodsFirstly, an electrically heated analog heat source was designed and imitated based on various actual GPHS performance parameters. Based on the simulation calculations, the thermodynamic equivalent substitutability between the simulated heat source and real GPHS in terms of material and dimensional differences was evaluated. Then, the performance of its operation in different application scenarios was analyzed, and an optimized application environment was proposed. Finally, based on the experimental test results, the uniformity of the thermal output characteristics of the imitation GPHS-simulated heat source was compared with that of a real GPHS, and the practical application characteristics of the imitation GPHS in RTGs were also evaluated.ResultsAt an input power of 250 W, the average surface temperature of the GPHS-simulated heat source reaches 515 K. The temperature variation trend with power is consistent with that of the simulation results. Experimental test results show that the energy conversion efficiency of the RTG module is increased to ~6% with a 250-W heat source power supply.ConclusionsThe proposed and constructed good equivalent simulated GPHS heat source with reference to the thermal properties of a real GPHS can be applied to RTGs and provides an effective and unified reference standard for the performance evaluation of radioisotope power sources.
- Published
- 2023
- Full Text
- View/download PDF
11. Evaluation of the determinants of food security within the COVID-19 pandemic circumstances- a particular case of Shaanxi, China
- Author
-
Apurbo Sarkar, Wang Hongyu, Abdul Azim Jony, Jiban Chandro Das, Waqar Hussain Memon, and Lu Qian
- Subjects
Food security ,Key-aspects ,Production ,Distribution ,Structural equation modeling ,Determinants ,Public aspects of medicine ,RA1-1270 - Abstract
Abstract Background Agricultural food production and distribution industries may play a vital role in determining the current conditions of any country’s food security and sustainable development goals. This paper examined the determinants of food security within three distinct aspects (effective utilization of food, food availability, and food access) within the COVID-19 epidemic situation. Methods The qualitative set-up of the study comprised with the identification of drivers by critical analysis of published papers and discussion held with some practitioners. The quantitative data used in this research were collected from a survey covering the agricultural food supply industry in China (Shaanxi Province). The survey was conducted from November to December 2020 and we mainly focus on three aspects of food security (effective utilization of food, food availability, and food access). The core analytical assumptions were made by employing exploratory factor analysis (EFA), confirmatory factor analysis (CFA), and structural equation modeling (SEM). Results After analyzing the data collected from 257 agricultural food productions and distribution personnel along with the hypothesis testing, it found that the determinants of the effective utilization of food were positively related to the determinants of food access (β = 0.291, p = 0.029) and food availability (β = 0.298, p = 0.011), and the determinants of food availability were positively related to the food access determinants (β = 0.128, p = 0.002). The association and variance values between food availability and food access were 0.659 and 0.407; the association and variance values between for effective utilization of food and food availability aspects were 0.465 and 0.298, and between effective utilization of food and economy were 0.508 and 0.475. Conclusion The study critically evaluated the interconnection among the crucial determinants within the banner of three dimensions, which will act as a major contribution to existing literature. This research will help the government and industry to develop policies and strategies for the successful implementation of all the associated determinants of food security in terms of the epidemic situation.
- Published
- 2021
- Full Text
- View/download PDF
12. Cross Fusion RGB-T Tracking with Bi-directional Adapter
- Author
-
Zeng, Zhirong, Liu, Xiaotao, Sun, Meng, Wang, Hongyu, and Liu, Jing
- Subjects
Computer Science - Computer Vision and Pattern Recognition - Abstract
Many state-of-the-art RGB-T trackers have achieved remarkable results through modality fusion. However, these trackers often either overlook temporal information or fail to fully utilize it, resulting in an ineffective balance between multi-modal and temporal information. To address this issue, we propose a novel Cross Fusion RGB-T Tracking architecture (CFBT) that ensures the full participation of multiple modalities in tracking while dynamically fusing temporal information. The effectiveness of CFBT relies on three newly designed cross spatio-temporal information fusion modules: Cross Spatio-Temporal Augmentation Fusion (CSTAF), Cross Spatio-Temporal Complementarity Fusion (CSTCF), and Dual-Stream Spatio-Temporal Adapter (DSTA). CSTAF employs a cross-attention mechanism to enhance the feature representation of the template comprehensively. CSTCF utilizes complementary information between different branches to enhance target features and suppress background features. DSTA adopts the adapter concept to adaptively fuse complementary information from multiple branches within the transformer layer, using the RGB modality as a medium. These ingenious fusions of multiple perspectives introduce only less than 0.3\% of the total modal parameters, but they indeed enable an efficient balance between multi-modal and temporal information. Extensive experiments on three popular RGB-T tracking benchmarks demonstrate that our method achieves new state-of-the-art performance.
- Published
- 2024
13. Enhancement of Co-located Shared VR Experiences: Representing Non-HMD Observers on Both HMD and 2D Screen
- Author
-
Guo, Zixuan, Xu, Wenge, Wang, Hongyu, Wan, Tingjie, Baghaei, Nilufar, Lo, Cheng-Hung, and Liang, Hai-Ning
- Subjects
Computer Science - Human-Computer Interaction - Abstract
Virtual reality (VR) not only allows head-mounted display (HMD) users to immerse themselves in virtual worlds but also to share them with others. When designed correctly, this shared experience can be enjoyable. However, in typical scenarios, HMD users are isolated by their devices, and non-HMD observers lack connection with the virtual world. To address this, our research investigates visually representing observers on both HMD and 2D screens to enhance shared experiences. The study, including five representation conditions, reveals that incorporating observer representation positively impacts both HMD users and observers. For how to design and represent them, our work shows that HMD users prefer methods displaying real-world visuals, while observers exhibit diverse preferences regarding being represented with real or virtual images. We provide design guidelines tailored to both displays, offering valuable insights to enhance co-located shared VR experiences for HMD users and non-HMD observers.
- Published
- 2024
14. Exploring the Impact of Passthrough on VR Exergaming in Public Environments: A Field Study
- Author
-
Guo, Zixuan, Deng, Hanxiao, Wang, Hongyu, Tan, Angel J. Y., Xu, Wenge, and Liang, Hai-Ning
- Subjects
Computer Science - Human-Computer Interaction - Abstract
Sedentary behavior is becoming increasingly prevalent in daily work and study environments. VR exergaming has emerged as a promising solution in these places of work and study. However, private spaces in these environments are not easy, and engaging in VR exergaming in public settings presents its own set of challenges (e.g., safety, social acceptance, isolation, and privacy protection). The recent development of Passthrough functionality in VR headsets allows users to maintain awareness of their surroundings, enhancing safety and convenience. Despite its potential benefits, little is known about how Passthrough could affect user performance and experience and solve the challenges of playing VR exergames in real-world public environments. To our knowledge, this work is the first to conduct a field study in an underground passageway on a university campus to explore the use of Passthrough in a real-world public environment, with a disturbance-free closed room as a baseline. Results indicate that enabling Passthrough in a public environment improves performance without compromising presence. Moreover, Passthrough can increase social acceptance, especially among individuals with higher levels of self-consciousness. These findings highlight Passthrough's potential to encourage VR exergaming adoption in public environments, with promising implications for overall health and well-being.
- Published
- 2024
15. Q-Sparse: All Large Language Models can be Fully Sparsely-Activated
- Author
-
Wang, Hongyu, Ma, Shuming, Wang, Ruiping, and Wei, Furu
- Subjects
Computer Science - Computation and Language ,Computer Science - Machine Learning - Abstract
We introduce, Q-Sparse, a simple yet effective approach to training sparsely-activated large language models (LLMs). Q-Sparse enables full sparsity of activations in LLMs which can bring significant efficiency gains in inference. This is achieved by applying top-K sparsification to the activations and the straight-through-estimator to the training. We also introduce Block Q-Sparse for batch training and inference. The key results from this work are, (1) Q-Sparse can achieve results comparable to those of baseline LLMs while being much more efficient at inference time; (2) We present an inference-optimal scaling law for sparsely-activated LLMs; (3) Q-Sparse is effective in different settings, including training-from-scratch, continue-training of off-the-shelf LLMs, and finetuning; (4) Q-Sparse works for both full-precision and 1-bit LLMs (e.g., BitNet b1.58). Particularly, the synergy of BitNet b1.58 and Q-Sparse (can be equipped with MoE) provides the cornerstone and a clear path to revolutionize the efficiency, including cost and energy consumption, of future LLMs., Comment: Work in progress
- Published
- 2024
16. HTD-Mamba: Efficient Hyperspectral Target Detection with Pyramid State Space Model
- Author
-
Shen, Dunbin, Zhu, Xuanbing, Tian, Jiacheng, Liu, Jianjun, Du, Zhenrong, Wang, Hongyu, and Ma, Xiaorui
- Subjects
Computer Science - Computer Vision and Pattern Recognition - Abstract
Hyperspectral target detection (HTD) identifies objects of interest from complex backgrounds at the pixel level, playing a vital role in Earth observation. However, HTD faces challenges due to limited prior knowledge and spectral variation, leading to underfitting models and unreliable performance. To address these challenges, this paper proposes an efficient self-supervised HTD method with a pyramid state space model (SSM), named HTD-Mamba, which employs spectrally contrastive learning to distinguish between target and background based on the similarity measurement of intrinsic features. Specifically, to obtain sufficient training samples and leverage spatial contextual information, we propose a spatial-encoded spectral augmentation technique that encodes all surrounding pixels within a patch into a transformed view of the center pixel. Additionally, to explore global band correlations, we divide pixels into continuous group-wise spectral embeddings and introduce Mamba to HTD for the first time to model long-range dependencies of the spectral sequence with linear complexity. Furthermore, to alleviate spectral variation and enhance robust representation, we propose a pyramid SSM as a backbone to capture and fuse multiresolution spectral-wise intrinsic features. Extensive experiments conducted on four public datasets demonstrate that the proposed method outperforms state-of-the-art methods in both quantitative and qualitative evaluations. Code is available at \url{https://github.com/shendb2022/HTD-Mamba}., Comment: 13 pages,6 figures, 5 tables
- Published
- 2024
17. Voltage control strategy for distribution network based on SoC FPGA hardware parallel computing
- Author
-
DANG Haotian, LIU Dong, CHEN Fei, ZHAO Xianping, LIU Siyang, and WANG Hongyu
- Subjects
active distribution network ,voltage control ,edge computing ,system on chip field programmable gate array (soc fpga) ,parallel computing ,genetic algorithm ,Applications of electric power ,TK4001-4102 - Abstract
With the development of active distribution network and Internet of Things technology,the access of reactive power equipment is becoming more complicated and marginalized,and the related computing of voltage control is also developing towards edge computing. Due to the limited computing power,the pure software calculation of edge nodes takes a long time,which cannot meet the requirements of real-time control. Aiming at this problem,a distribution network voltage control strategy based on system on chip field programmable gate array (SoC FPGA) hardware parallel computing is proposed. Firstly,the software and hardware computing framework based on SoC FPGA is designed. Then,targeted improvements are made to the distribution network voltage control model and genetic algorithm solution method for FPGA computing. Finally,the FPGA hardware solution structure is designed in modules. The verification of example scenarios shows that,compared with the pure software solution method of edge nodes,the average solution efficiency of the proposed strategy in scenarios where voltage is lower than lower limit and where voltage exceeds upper limit is increased by 2.41 times and 2.15 times respectively,which can effectively improve the real-time performance of voltage control.
- Published
- 2022
- Full Text
- View/download PDF
18. Development of a Hybrid Solar and Wind-Powered Long-Range Unmanned Ocean Stereo Exploration Vessel
- Author
-
YAO Tiancheng, ZHAO Yongsheng, WANG Hongyu, HE Yanping, DING Zilong, CHI Zheying, and CAI Weikai
- Subjects
unmanned surface vessel (usv) ,long duration ,ship borne tethered unmanned aerial vehicle (uav) ,ocean-air stereo exploration ,ducted wind turbine ,Engineering (General). Civil engineering (General) ,TA1-2040 ,Chemical engineering ,TP155-156 ,Naval architecture. Shipbuilding. Marine engineering ,VM1-989 - Abstract
Aimed at the problems of unmanned marine exploration vessels, such as the short voyage time and the limited sensing ability caused by sensor failure under complex marine environments, a long-range unmanned ocean-air stereo exploration vessel driven by wind and solar energy is developed. An elevating ducted wind turbine is designed for high efficiency and low starting wind speed, and a deployable solar photovoltaic power generation system is developed. Therefore, wind power and solar energy can be utilized to realize a hybrid system, which overcomes the instability of single energy power supply, and effectively ensures the endurance of unmanned exploration vessel. Then, a ship-borne tethered ummanned aerial vehicle (UAV) system is developed with an autonomous takeoff and landing control section. Finally, the information fusion technology of ship borne and airborne sensors is adopted to greatly improve the perception ability of the unmanned ship to the surrounding environment and the function of three-dimension detection of sea and air. The unmanned surface vessel (USV) proposed in this paper is permitted to perform the assigned task with different types of loading equipment according to the scenarios.
- Published
- 2021
- Full Text
- View/download PDF
19. Research on Power IoT System Based City Block Air Pollutant Emission Prediction
- Author
-
Liu Siyang, Nie Yongjie, Li Bo, Zhu Mengyao, Li Zhengxing, Li Ting, Cao Min, Chang Yanping, Xu Hua, Yan Hongfeng, Jin Hui, and Wang Hongyu
- Subjects
Environmental sciences ,GE1-350 - Abstract
In order to constantly improve city environmental air quality, it is necessary to accurately control the major pollutants emissions such as air fine particulate matter. By adopting the proposed iterative update framework of air pollutant emission inventory, combined with block-level real-time electricity consumption data acquired by the smart city power IoT, and utilizing station-level and hourly environmental air quality monitoring data in specific areas of Yuxi and Dali in Yunnan Province from 2020 to 2021, the iterative update of emission inventory and prediction of air pollutant emission are studied. The experimental results shows that the prediction of the monthly average major air pollutants emissions such as NO2/PM10/PM2.5 in specific neighbourhoods of the two cities mentioned above reaches the same accuracy level as using numerical simulation prediction methods, but the prediction computational power requirements are greatly reduced, making it more suitable for the application requirements of the power IoT. This study provides a new idea for improving the regulatory capacity of intelligent environment and achieving higher urban air quality based on the smart city power IoT.
- Published
- 2023
- Full Text
- View/download PDF
20. M4U: Evaluating Multilingual Understanding and Reasoning for Large Multimodal Models
- Author
-
Wang, Hongyu, Xu, Jiayu, Xie, Senwei, Wang, Ruiping, Li, Jialin, Xie, Zhaojie, Zhang, Bin, Xiong, Chuyan, and Chen, Xilin
- Subjects
Computer Science - Computer Vision and Pattern Recognition ,Computer Science - Computation and Language - Abstract
Multilingual multimodal reasoning is a core component in achieving human-level intelligence. However, most existing benchmarks for multilingual multimodal reasoning struggle to differentiate between models of varying performance; even language models without visual capabilities can easily achieve high scores. This leaves a comprehensive evaluation of leading multilingual multimodal models largely unexplored. In this work, we introduce M4U, a novel and challenging benchmark for assessing the capability of multi-discipline multilingual multimodal understanding and reasoning. M4U contains 8,931 samples covering 64 disciplines across 16 subfields in Science, Engineering, and Healthcare in Chinese, English, and German. Using M4U, we conduct extensive evaluations of 21 leading Large Multimodal Models (LMMs) and Large Language Models (LLMs) with external tools. The evaluation results show that the state-of-the-art model, GPT-4o, achieves only 47.6% average accuracy on M4U. Additionally, we observe that the leading LMMs exhibit significant language preferences. Our in-depth analysis indicates that leading LMMs, including GPT-4o, suffer performance degradation when prompted with cross-lingual multimodal questions, such as images with key textual information in Chinese while the question is in German. We believe that M4U can serve as a crucial tool for systematically evaluating LMMs based on their multilingual multimodal reasoning capabilities and monitoring their development. The homepage, codes and data are public available., Comment: Work in progress
- Published
- 2024
21. Real-Time and Accurate: Zero-shot High-Fidelity Singing Voice Conversion with Multi-Condition Flow Synthesis
- Author
-
Li, Hui, Wang, Hongyu, Chen, Zhijin, Sun, Bohan, and Li, Bo
- Subjects
Electrical Engineering and Systems Science - Audio and Speech Processing - Abstract
Singing voice conversion is to convert the source sing voice into the target sing voice except for the content. Currently, flow-based models can complete the task of voice conversion, but they struggle to effectively extract latent variables in the more rhythmically rich and emotionally expressive task of singing voice conversion, while also facing issues with low efficiency in speech processing. In this paper, we propose a high-fidelity flow-based model based on multi-decoupling feature constraints, which enhances the capture of vocal details by integrating multiple encoders. We also use iSTFT to enhance the speed of speech processing by replacing some layers of the Vocoder. We compare the synthesized singing voice with other models from multiple dimensions, and our proposed model is highly consistent with the current state-of-the-art, with the demo which is available at \url{https://lazycat1119.github.io/RASVC-demo/}, Comment: 5 pages,4 figures
- Published
- 2024
22. NGM-SLAM: Gaussian Splatting SLAM with Radiance Field Submap
- Author
-
Li, Mingrui, Huang, Jingwei, Sun, Lei, Tian, Aaron Xuxiang, Deng, Tianchen, and Wang, Hongyu
- Subjects
Computer Science - Robotics - Abstract
SLAM systems based on Gaussian Splatting have garnered attention due to their capabilities for rapid real-time rendering and high-fidelity mapping. However, current Gaussian Splatting SLAM systems usually struggle with large scene representation and lack effective loop closure detection. To address these issues, we introduce NGM-SLAM, the first 3DGS based SLAM system that utilizes neural radiance field submaps for progressive scene expression, effectively integrating the strengths of neural radiance fields and 3D Gaussian Splatting. We utilize neural radiance field submaps as supervision and achieve high-quality scene expression and online loop closure adjustments through Gaussian rendering of fused submaps. Our results on multiple real-world scenes and large-scale scene datasets demonstrate that our method can achieve accurate hole filling and high-quality scene expression, supporting monocular, stereo, and RGB-D inputs, and achieving state-of-the-art scene reconstruction and tracking performance., Comment: 9pages, 4 figures
- Published
- 2024
23. PCG: Mitigating Conflict-based Cache Side-channel Attacks with Prefetching
- Author
-
Jiang, Fang, Tong, Fei, Wang, Hongyu, Cheng, Xiaoyu, Zhou, Zhe, Ling, Ming, and Mao, Yuxing
- Subjects
Computer Science - Cryptography and Security ,Computer Science - Hardware Architecture - Abstract
To defend against conflict-based cache side-channel attacks, cache partitioning or remapping techniques were proposed to prevent set conflicts between different security domains or obfuscate the locations of such conflicts. But such techniques complicate cache design and may result in significant performance penalties. Therefore, there have been lightweight prefetching-based schemes proposed to introduce noise to confuse attackers' observation. However, we have validated experimentally that relying on prefetching to only introduce noise is insufficient, as attackers can still reliably distinguish the victim's cache accesses. This paper proposes a novel prefetching-based scheme, called PCG. It combines adding victim-irrelevant cache occupancy changes and reducing victim-relevant cache occupancy changes to disrupt attackers by generating noisy and indistinguishable cache access patterns. Additionally, PCG can either work independently or seamlessly be integrated with most of the commonly used prefetchers. We have implemented and evaluated PCG in both gem5 and the open-source RISC-V core BOOMv3. The evaluation results show the PCG's robust security superior to the existing solutions, while without resulting in significant performance degradation. According to the evaluation based on the SPEC CPU 2017 benchmark suite, PCG even shows an average performance improvement of about 1.64%. Moreover, it incurs only 1.26% overhead on hardware resource consumption., Comment: 12 pages, 9 figures, submitting to a journal
- Published
- 2024
24. Prompt-Guided Generation of Structured Chest X-Ray Report Using a Pre-trained LLM
- Author
-
Li, Hongzhao, Wang, Hongyu, Sun, Xia, He, Hua, and Feng, Jun
- Subjects
Computer Science - Artificial Intelligence ,Computer Science - Computer Vision and Pattern Recognition ,Computer Science - Multimedia - Abstract
Medical report generation automates radiology descriptions from images, easing the burden on physicians and minimizing errors. However, current methods lack structured outputs and physician interactivity for clear, clinically relevant reports. Our method introduces a prompt-guided approach to generate structured chest X-ray reports using a pre-trained large language model (LLM). First, we identify anatomical regions in chest X-rays to generate focused sentences that center on key visual elements, thereby establishing a structured report foundation with anatomy-based sentences. We also convert the detected anatomy into textual prompts conveying anatomical comprehension to the LLM. Additionally, the clinical context prompts guide the LLM to emphasize interactivity and clinical requirements. By integrating anatomy-focused sentences and anatomy/clinical prompts, the pre-trained LLM can generate structured chest X-ray reports tailored to prompted anatomical regions and clinical contexts. We evaluate using language generation and clinical effectiveness metrics, demonstrating strong performance., Comment: Accepted by IEEE Conference on Multimedia Expo 2024
- Published
- 2024
25. Illicit Promotion on Twitter
- Author
-
Wang, Hongyu, Li, Ying, Huang, Ronghong, and Mi, Xianghang
- Subjects
Computer Science - Cryptography and Security ,Computer Science - Social and Information Networks - Abstract
In this paper, we present an extensive study of the promotion of illicit goods and services on Twitter, a popular online social network(OSN). This study is made possible through the design and implementation of multiple novel tools for detecting and analyzing illicit promotion activities as well as their underlying campaigns. As the results, we observe that illicit promotion is prevalent on Twitter, along with noticeable existence on other three popular OSNs including Youtube, Facebook, and TikTok. Particularly, 12 million distinct posts of illicit promotion (PIPs) have been observed on the Twitter platform, which are widely distributed in 5 major natural languages and 10 categories of illicit goods and services, e.g., drugs, data leakage, gambling, and weapon sales. What are also observed are 580K Twitter accounts publishing PIPs as well as 37K distinct instant messaging (IM) accounts that are embedded in PIPs and serve as next hops of communication, which strongly indicates that the campaigns underpinning PIPs are also of a large scale. Also, an arms race between Twitter and illicit promotion operators is also observed. On one hand, Twitter is observed to conduct content moderation in a continuous manner and almost 80% PIPs will get gradually unpublished within six months since posted. However, in the meantime, miscreants adopt various evasion tactics to masquerade their PIPs, which renders more than 90% PIPs keeping hidden from the detection radar for two months or longer.
- Published
- 2024
26. The Era of 1-bit LLMs: All Large Language Models are in 1.58 Bits
- Author
-
Ma, Shuming, Wang, Hongyu, Ma, Lingxiao, Wang, Lei, Wang, Wenhui, Huang, Shaohan, Dong, Li, Wang, Ruiping, Xue, Jilong, and Wei, Furu
- Subjects
Computer Science - Computation and Language ,Computer Science - Machine Learning - Abstract
Recent research, such as BitNet, is paving the way for a new era of 1-bit Large Language Models (LLMs). In this work, we introduce a 1-bit LLM variant, namely BitNet b1.58, in which every single parameter (or weight) of the LLM is ternary {-1, 0, 1}. It matches the full-precision (i.e., FP16 or BF16) Transformer LLM with the same model size and training tokens in terms of both perplexity and end-task performance, while being significantly more cost-effective in terms of latency, memory, throughput, and energy consumption. More profoundly, the 1.58-bit LLM defines a new scaling law and recipe for training new generations of LLMs that are both high-performance and cost-effective. Furthermore, it enables a new computation paradigm and opens the door for designing specific hardware optimized for 1-bit LLMs., Comment: Work in progress
- Published
- 2024
27. SGS-SLAM: Semantic Gaussian Splatting For Neural Dense SLAM
- Author
-
Li, Mingrui, Liu, Shuhong, Zhou, Heng, Zhu, Guohao, Cheng, Na, Deng, Tianchen, and Wang, Hongyu
- Subjects
Computer Science - Computer Vision and Pattern Recognition ,Computer Science - Artificial Intelligence ,Computer Science - Robotics - Abstract
We present SGS-SLAM, the first semantic visual SLAM system based on Gaussian Splatting. It incorporates appearance, geometry, and semantic features through multi-channel optimization, addressing the oversmoothing limitations of neural implicit SLAM systems in high-quality rendering, scene understanding, and object-level geometry. We introduce a unique semantic feature loss that effectively compensates for the shortcomings of traditional depth and color losses in object optimization. Through a semantic-guided keyframe selection strategy, we prevent erroneous reconstructions caused by cumulative errors. Extensive experiments demonstrate that SGS-SLAM delivers state-of-the-art performance in camera pose estimation, map reconstruction, precise semantic segmentation, and object-level geometric accuracy, while ensuring real-time rendering capabilities.
- Published
- 2024
28. DDN-SLAM: Real-time Dense Dynamic Neural Implicit SLAM
- Author
-
Li, Mingrui, Zhou, Yiming, Jiang, Guangan, Deng, Tianchen, Wang, Yangyang, and Wang, Hongyu
- Subjects
Computer Science - Computer Vision and Pattern Recognition ,Computer Science - Robotics - Abstract
SLAM systems based on NeRF have demonstrated superior performance in rendering quality and scene reconstruction for static environments compared to traditional dense SLAM. However, they encounter tracking drift and mapping errors in real-world scenarios with dynamic interferences. To address these issues, we introduce DDN-SLAM, the first real-time dense dynamic neural implicit SLAM system integrating semantic features. To address dynamic tracking interferences, we propose a feature point segmentation method that combines semantic features with a mixed Gaussian distribution model. To avoid incorrect background removal, we propose a mapping strategy based on sparse point cloud sampling and background restoration. We propose a dynamic semantic loss to eliminate dynamic occlusions. Experimental results demonstrate that DDN-SLAM is capable of robustly tracking and producing high-quality reconstructions in dynamic environments, while appropriately preserving potential dynamic objects. Compared to existing neural implicit SLAM systems, the tracking results on dynamic datasets indicate an average 90% improvement in Average Trajectory Error (ATE) accuracy., Comment: 11pages, 4figures
- Published
- 2024
29. Temporal Adaptive RGBT Tracking with Modality Prompt
- Author
-
Wang, Hongyu, Liu, Xiaotao, Li, Yifan, Sun, Meng, Yuan, Dian, and Liu, Jing
- Subjects
Computer Science - Computer Vision and Pattern Recognition - Abstract
RGBT tracking has been widely used in various fields such as robotics, surveillance processing, and autonomous driving. Existing RGBT trackers fully explore the spatial information between the template and the search region and locate the target based on the appearance matching results. However, these RGBT trackers have very limited exploitation of temporal information, either ignoring temporal information or exploiting it through online sampling and training. The former struggles to cope with the object state changes, while the latter neglects the correlation between spatial and temporal information. To alleviate these limitations, we propose a novel Temporal Adaptive RGBT Tracking framework, named as TATrack. TATrack has a spatio-temporal two-stream structure and captures temporal information by an online updated template, where the two-stream structure refers to the multi-modal feature extraction and cross-modal interaction for the initial template and the online update template respectively. TATrack contributes to comprehensively exploit spatio-temporal information and multi-modal information for target localization. In addition, we design a spatio-temporal interaction (STI) mechanism that bridges two branches and enables cross-modal interaction to span longer time scales. Extensive experiments on three popular RGBT tracking benchmarks show that our method achieves state-of-the-art performance, while running at real-time speed.
- Published
- 2024
30. PLE-SLAM: A Visual-Inertial SLAM Based on Point-Line Features and Efficient IMU Initialization
- Author
-
He, Jiaming, Li, Mingrui, Wang, Yangyang, and Wang, Hongyu
- Subjects
Computer Science - Robotics - Abstract
Visual-inertial SLAM is crucial in various fields, such as aerial vehicles, industrial robots, and autonomous driving. The fusion of camera and inertial measurement unit (IMU) makes up for the shortcomings of a signal sensor, which significantly improves the accuracy and robustness of localization in challenging environments. This article presents PLE-SLAM, an accurate and real-time visual-inertial SLAM algorithm based on point-line features and efficient IMU initialization. First, we use parallel computing methods to extract features and compute descriptors to ensure real-time performance. Adjacent short line segments are merged into long line segments, and isolated short line segments are directly deleted. Second, a rotation-translation-decoupled initialization method is extended to use both points and lines. Gyroscope bias is optimized by tightly coupling IMU measurements and image observations. Accelerometer bias and gravity direction are solved by an analytical method for efficiency. To improve the system's intelligence in handling complex environments, a scheme of leveraging semantic information and geometric constraints to eliminate dynamic features and A solution for loop detection and closed-loop frame pose estimation using CNN and GNN are integrated into the system. All networks are accelerated to ensure real-time performance. The experiment results on public datasets illustrate that PLE-SLAM is one of the state-of-the-art visual-inertial SLAM systems.
- Published
- 2024
31. Eco-efficient recycling of engineering muck for manufacturing low-carbon geopolymers assessed through LCA: exploring the impact of synthesis conditions on performance
- Author
-
Yuan, Bingxiang, Liang, Jingkang, Huang, Xianlun, Huang, Qingyu, Zhang, Baifa, Yang, Guanghua, Wang, Yonghong, Yuan, Junhong, Wang, Hongyu, and Yuan, Peng
- Published
- 2024
- Full Text
- View/download PDF
32. Holes position prediction and inverse design on complex surface in deep-drawing process with sand dies based on NURBS and deformation mathematical zoning
- Author
-
Li, Yuanyuan, Wang, Hongyu, Liu, Menghan, Yang, Dehui, Sun, Jie, Zhang, Shunhu, and Ma, Xiangkun
- Published
- 2024
- Full Text
- View/download PDF
33. Identification of carbonates with high positive carbon isotope excursion from the Liaohe Group in the northeastern North China Craton and implications for the Lomagundi-Jatuli Event
- Author
-
Wang, Hongyu, Zhang, Shuanhong, Wang, Sen, Kong, Linghao, Zhao, Yue, Zhang, Qiqi, Gao, Sen, and Hu, Guohui
- Published
- 2024
- Full Text
- View/download PDF
34. Microstructural Analysis on Shear Behavior of New-fill and Silt Interface in Check Dam
- Author
-
Wang, Ya, Wang, Hongyu, and Guo, Liping
- Published
- 2024
- Full Text
- View/download PDF
35. Output-Space Outer Approximation Branch-and-Bound Algorithm for a Class of Linear Multiplicative Programs
- Author
-
Zhang, Bo, Wang, Hongyu, and Gao, Yuelin
- Published
- 2024
- Full Text
- View/download PDF
36. Combined brain topological metrics with machine learning to distinguish essential tremor and tremor-dominant Parkinson’s disease
- Author
-
Xiao, Pan, Li, Qin, Gui, Honge, Xu, Bintao, Zhao, Xiaole, Wang, Hongyu, Tao, Li, Chen, Huiyue, Wang, Hansheng, Lv, Fajin, Luo, Tianyou, Cheng, Oumei, Luo, Jin, Man, Yun, Xiao, Zheng, and Fang, Weidong
- Published
- 2024
- Full Text
- View/download PDF
37. Analysis and simulation research on deep drawing of magnesium alloy box-shaped products with sand die based on different driving modes
- Author
-
Han, Xinyu, Wang, Hongyu, Wang, Xinyu, Han, Jinyan, and Zhang, Sixiang
- Published
- 2024
- Full Text
- View/download PDF
38. Non-reciprocal wave propagations in a one-dimensional periodic structure modified with a linkage mechanism
- Author
-
Wang, Hongyu, Zhao, Jian, Wang, Xuefeng, Dong, Zeyuan, and Huang, Yu
- Published
- 2024
- Full Text
- View/download PDF
39. Improvement for tolerance of lithium metal counter electrodes towards sodium contamination in hybrid Li/Na-ion electrolytes
- Author
-
Liu, Liyang, Zhao, Xufan, Qi, Jiaxing, Abdussalam, Abubakar, Zhang, Wei, Wang, Hongyu, and Xu, Guobao
- Published
- 2024
- Full Text
- View/download PDF
40. Insight into the competitiveness of C–C coupling reactions of 5-hydroxymethylfurfural with lignocellulosic compounds in one pot
- Author
-
Song, Linyan, Nie, Genkuo, Chen, Xiulei, Wang, Hongyu, Liu, Shiwei, Yu, Hailong, Liu, Xuguang, Yang, Guihua, and Yu, Shitao
- Published
- 2024
- Full Text
- View/download PDF
41. Research on automatic pavement crack identification Based on improved YOLOv8
- Author
-
Wang, Hongyu, Han, Xiao, Song, Xifa, Su, Jie, Li, Yang, Zheng, Wenyan, and Wu, Xuejing
- Published
- 2024
- Full Text
- View/download PDF
42. Printable oil-NiOX hole transport layer enables efficient n-i-p perovskite solar cells with a high thermal stability
- Author
-
Xu, Yazhou, Chu, Zhaoyang, Wang, Hongyu, Xing, Zhi, Fan, Baojin, Hu, Xiaotian, and Chen, Yiwang
- Published
- 2024
- Full Text
- View/download PDF
43. Multiple hypergraph convolutional network social recommendation using dual contrastive learning
- Author
-
Wang, Hongyu, Zhou, Wei, Wen, Junhao, and Qiao, Shutong
- Published
- 2024
- Full Text
- View/download PDF
44. Chronic GLP1 therapy reduces postprandial IL6 in obese humans with prediabetes.
- Author
-
Hamidi, Vala, Wang, Hongyu, Pham, Vi, Bermudez Saint Andre, Karla, Taegtmeyer, Heinrich, and Gutierrez, Absalon
- Subjects
GLP1 ,IL6 ,inflammation ,obesity ,prediabetes - Abstract
Single-dose glucagon-like peptide 1 (GLP1) therapy increases postprandial plasma IL6 levels in prediabetic, obese humans. GLP1-IL6 interactions underly multiple antidiabetic effects, but these may differ after acute versus chronic therapy. This study examines postprandial effects of GLP1 after chronic therapy. Seven humans (six Black) with prediabetes and obesity completed 6 weeks of exenatide extended release therapy. Then subjects returned for pre- and post-meal measurements of plasma IL6, GLP1, glucagon, and related inflammatory markers. Weight, which was measured before and after therapy, did not change. Plasma IL6 decreased from baseline to postmeal state ( = 0.016), with decreases in free fatty acids (P
- Published
- 2024
45. VOT: Revolutionizing Speaker Verification with Memory and Attention Mechanisms
- Author
-
Wang, Hongyu, Li, Hui, and Li, Bo
- Subjects
Electrical Engineering and Systems Science - Audio and Speech Processing ,Computer Science - Sound - Abstract
Speaker verification is to judge the similarity of two unknown voices in an open set, where the ideal speaker embedding should be able to condense discriminant information into a compact utterance-level representation that has small intra-speaker distances and large inter-speaker distances.We propose a novel model named Voice Transformer(VOT) for speaker verification. The model consists of multiple parallel Transformers, and the outputs of these Transformers are adaptively combined. Deeply-Fused Semantic Memory Network(DFSMN)is integrated into the attention parts of these Transformers to capture long-distance information and enhance the local dependencies. Statistical pooling layers are incorporated to enhance overall performance without significantly increasing the number of parameters. We propose a new loss function called Additive Angular Margin Focal Loss(AAMF) to address the hard sample mining issue.We evaluate the proposed approach on the VoxCeleb1 and CN-Celeb2 datasets. The experimental results demonstrate that VOT achieves state-of-the-art results, outperforming nearly all existing models. The code is available on GitHub., Comment: 8 pages,4 figures,6 tables
- Published
- 2023
46. Who's Watching Me?: Exploring the Impact of Audience Familiarity on Player Performance, Experience, and Exertion in Virtual Reality Exergames
- Author
-
Guo, Zixuan, Xu, Wenge, Zhang, Jialin, Wang, Hongyu, Lo, Cheng-Hung, and Liang, Hai-Ning
- Subjects
Computer Science - Human-Computer Interaction - Abstract
Familiarity with audiences plays a significant role in shaping individual performance and experience across various activities in everyday life. This study delves into the impact of familiarity with non-playable character (NPC) audiences on player performance and experience in virtual reality (VR) exergames. By manipulating of NPC appearance (face and body shape) and voice familiarity, we explored their effect on game performance, experience, and exertion. The findings reveal that familiar NPC audiences have a positive impact on performance, creating a more enjoyable gaming experience, and leading players to perceive less exertion. Moreover, individuals with higher levels of self-consciousness exhibit heightened sensitivity to the familiarity with NPC audiences. Our results shed light on the role of familiar NPC audiences in enhancing player experiences and provide insights for designing more engaging and personalized VR exergame environments., Comment: 10 pages, 5 figures, IEEE International Symposium on Mixed and Augmented Reality (ISMAR) 2023
- Published
- 2023
47. BitNet: Scaling 1-bit Transformers for Large Language Models
- Author
-
Wang, Hongyu, Ma, Shuming, Dong, Li, Huang, Shaohan, Wang, Huaijie, Ma, Lingxiao, Yang, Fan, Wang, Ruiping, Wu, Yi, and Wei, Furu
- Subjects
Computer Science - Computation and Language - Abstract
The increasing size of large language models has posed challenges for deployment and raised concerns about environmental impact due to high energy consumption. In this work, we introduce BitNet, a scalable and stable 1-bit Transformer architecture designed for large language models. Specifically, we introduce BitLinear as a drop-in replacement of the nn.Linear layer in order to train 1-bit weights from scratch. Experimental results on language modeling show that BitNet achieves competitive performance while substantially reducing memory footprint and energy consumption, compared to state-of-the-art 8-bit quantization methods and FP16 Transformer baselines. Furthermore, BitNet exhibits a scaling law akin to full-precision Transformers, suggesting its potential for effective scaling to even larger language models while maintaining efficiency and performance benefits., Comment: Work in progress
- Published
- 2023
48. The Gehring-Hayman type theorem on pseudoconvex domains of finite type in $\mathbb{C}^2$
- Author
-
Li, Haichou, Pu, Xingsi, and Wang, Hongyu
- Subjects
Mathematics - Complex Variables - Abstract
In this paper, we obtain the Gehring-Hayman type theorem on smoothly bounded pseudoconvex domains of finite type in $\mathbb{C}^2$. As an application, we provide a quantitative comparison between global and local Kobayashi distances near a boundary point for these domains., Comment: arXiv admin note: substantial text overlap with arXiv:2301.06411
- Published
- 2023
49. Electrical Characteristics of the GEC Reference Cell with Impedance Matching: A Two-Dimensional PIC/MCC Modeling Study
- Author
-
Chen, Zili, Wang, Hongyu, Yu, Shimin, Wang, Yu, Chen, Zhipeng, Jiang, Wei, Schulze, Julian, and Zhang, Ya
- Subjects
Physics - Plasma Physics - Abstract
In this paper, the electrical characteristics of the Gaseous Electronics Conference (GEC) reference cell with impedance matching are investigated through a two-dimensional electrostatic implicit Particle-in-Cell/Monte Carlo Collision (PIC/MCC) model in an axisymmetric coordinate system. The coupling between the complex reactor geometry and the external circuit is included via an equivalent capacitance calculated from the electric energy density. The results of this model are compared with experimental measurements and other model calculations and show good agreement. This simulation obtains the plasma kinetics of the capacitively coupled discharge process at low pressure and detailed external circuit responses, including power transmission, reflection, and higher-order harmonics in the circuit, which provides important insights for impedance-matching design in semiconductor plasma processing.
- Published
- 2023
50. Analysis of non-physiological shear stress-induced red blood cell trauma across different clinical support conditions of the blood pump
- Author
-
Liu, Xinyu, Li, Yuan, Jia, Jinze, Wang, Hongyu, Xi, Yifeng, Sun, Anqiang, Wang, Lizhen, Deng, Xiaoyan, Chen, Zengsheng, and Fan, Yubo
- 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.