检索结果-内蒙古大学图书馆

arXiv 2020年

作者： Chen, Ze Fu, Zhihang Jiang, Rongxin Chen, Yaowu Hua, Xian-sheng Zhejiang University Institute of Advanced Digital Technology and Instrument Zhejiang University Embedded System Engineering Research Center Ministry of Education of China Zhejiang University State Key Laboratory of Industrial Control Technology Zhejiang Provincial Key Laboratory for Network Multimedia Technologies Alibaba DAMO Academy Alibaba Group

Based on the framework of multiple instance learning (MIL), tremendous works have promoted the advances of weakly supervised object detection (WSOD). However, most MIL-based methods tend to localize instances to their discriminative parts instead of the whole content. In this paper, we propose a spatial likelihood voting (SLV) module to converge the proposal localizing process without any bounding box annotations. Specifically, all region proposals in a given image play the role of voters every iteration during training, voting for the likelihood of each category in spatial dimensions. After dilating alignment on the area with large likelihood values, the voting results are regularized as bounding boxes, being used for the final classification and localization. Based on SLV, we further propose an end-to-end training framework for multi-task learning. The classification and localization tasks promote each other, which further improves the detection performance. Extensive experiments on the PASCAL VOC 2007 and 2012 datasets demonstrate the superior performance of SLV. Copyright © 2020, The Authors. All rights reserved.

关键词： Object recognition

来源：评论

学校读者我要写书评

暂无评论

Corrigendum: Terahertz spectral unmixing based method for identifying gastric cancer (2018 Phys. Med. Biol. 63 035016)

引用

Physics in Medicine & Biology 2022年第21期67卷 219501-219501页

作者： Yuqi Cao Pingjie Huang Xian Li Weiting Ge Dibo Hou Guangxin Zhang State Key Laboratory of Industrial Control Technology College of Control Science and Engineering Zhejiang University Hangzhou People’s Republic of China Cancer Institute (Key Laboratory of Cancer Prevention and Intervention China National Ministry of Education Key Laboratory of Molecular Biology in Medical Sciences Zhejiang Province China) The Second Affiliated Hospital Zhejiang University School of Medicine Hangzhou People’s Republic of China

来源：评论

学校读者我要写书评

暂无评论

Multi-task GANs for semantic segmentation and depth completion with cycle consistency

arXiv

引用

arXiv 2020年

作者： Zhang, Chongzhen Tang, Yang Zhao, Chaoqiang Sun, Qiyu Ye, Zhencheng Kurths, Jürgen Key Laboratory of Advanced Control and Optimization for Chemical Process Ministry of Education East China University of Science and Technology Shanghai200237 China Potsdam Institute for Climate Impact Research Potsdam14473 Germany Institute of Physics Humboldt University of Berlin Berlin12489 Germany

Semantic segmentation and depth completion are two challenging tasks in scene understanding, and they are widely used in robotics and autonomous driving. Although several studies have been proposed to jointly train these two tasks using some small modifications, like changing the last layer, the result of one task is not utilized to improve the performance of the other one despite that there are some similarities between these two tasks. In this paper, we propose multi-task generative adversarial networks (Multi-task GANs), which are not only competent in semantic segmentation and depth completion, but also improve the accuracy of depth completion through generated semantic images. In addition, we improve the details of generated semantic images based on CycleGAN by introducing multi-scale spatial pooling blocks and the structural similarity reconstruction loss. Furthermore, considering the inner consistency between semantic and geometric structures, we develop a semantic-guided smoothness loss to improve depth completion results. Extensive experiments on Cityscapes dataset and KITTI depth completion benchmark show that the Multi-task GANs are capable of achieving competitive performance for both semantic segmentation and depth completion tasks. © 2020, CC BY.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

A promising perspective to boost the utilizability of oil bodies: Moderate regulation and modification of interface

引用

Comprehensive reviews in food science and food safety 2025年第3期24卷 e70145页

作者： Shan Zhang David Julian McClements Ruofan Zheng Xiao Yu Zhida Sun Bijun Xie Yashu Chen Qianchun Deng Oil Crops Research Institute of the Chinese Academy of Agricultural Sciences Oil Crops and Lipids Process Technology National & Local Joint Engineering Laboratory Key Laboratory of Oilseeds Processing Ministry of Agriculture Hubei Key Laboratory of Lipid Chemistry and Nutrition Wuhan China. Natural Product Laboratory Department of Food Science and Technology Huazhong Agricultural University Wuhan Hubei China. Department of Food Science University of Massachusetts Amherst Massachusetts USA. College of Food and Bioengineering Zhengzhou University of Light Industry Henan Key Laboratory of Cold Chain Food Quality and Safety Control Collaborative Innovation Center for Food Production and Safety Zhengzhou Henan Province China.

Oil bodies (OBs), as natural oil storage organelles, can be obtained and utilized directly by simple extraction, thereby satisfying consumer demand for plant-derived green foods. The unique topological structure of the phospholipid-protein membrane renders OBs controllable and resistant to environmental stres, demonstrating their promising prospects for applications. As natural self-assemblies, the nutrient distribution, composition, and structural characteristics of OBs can be modified in multiple ways, including the genetic improvement and various processing techniques, to meet diverse application requirements. Generally, the utilization of OBs (e.g., processing stability, emulsifying, digestion, etc.) is closely related to their interfacial properties, but the associated studies have not been systematically reviewed. In this paper, we systematically review the structural and interfacial characteristics of OBs for the first time, encompassing their biosynthetic pathways and the structure-function relationship critical to their processability and bioavailability In particular, targeted improvement methods were also discussed. The underlying mechanisms of the physicochemical stability of OBs were primarily related to the interfacial modulation, including linkages between density, charge, and the number of protein-phospholipid salt bridges, as well as the quantity and structure of extrinsic proteins. The membrane compactness, enzyme binding sites, and aggregation of OBs in gastrointestinal tract significantly impact their digestion and subsequent metabolic fate. In summary, moderate interfacial modification, by altering the interactions between membrane components and retaining some extrinsic proteins, may be a promising approach to boost the stability and functionality of OBs.

关键词： gastrointestinal tract interfacial structure oil body physicochemical stability

来源：评论

学校读者我要写书评

暂无评论

Batch process monitoring based on WGNPE–GSVDD related and independent variables

引用

Chinese Journal of Chemical Engineering 2018年第12期26卷 2549-2561页

作者： Yongyong Hui Xiaoqiang Zhao College of Electrical and Information Engineering Lanzhou University of Technology Lanzhou 730050China Key Laboratory of Gansu Advanced Control for Industrial Processes Lanzhou University of TechnologyLanzhou 730050China National Experimental Teaching Center of Electrical and Control Engineering Lanzhou University of TechnologyLanzhou 730050China

In many batch processes, there are related or independence relationships among process variables. The traditional monitoring method usually carries out a single statistical model according to the related or independent method, and in the feature extraction there is not fully taken into account the characterization of fault information, it will make the process monitoring ineffective, so a fault monitoring method based on WGNPE(weighted global neighborhood preserving embedding)–GSVDD(greedy support vector data description) related and independent variables is proposed. First, mutual information method is used to separate the related variables and independent variables. Secondly, WGNPE method is used to extract the local and global structures of the related variables in batch process and highlight the fault information, GSVDD method is used to extract the process information of the independent variables quickly and effectively. Finally, the statistical monitoring model is established to achieve process monitoring based on WGNPE and GSVDD. The effectiveness of the proposed method was verified by the penicillin fermentation process.

关键词： Batch process Monitoring Related and independent variables Global-local Support vector data description

来源：评论

学校读者我要写书评

暂无评论

Effect of quorum-sensing molecule 2-phenylethanol on fermentation performance of Saccharomyces cerevisiae 31 in simulated sour fish

引用

Food Microbiology 2025年 132卷

作者： Pei Gao Shuhan Yang Qixing Jiang Peipei Yu Fang Yang Xiaojing Zhang Zhiqing Zhang Shaoquan Liu Wenshui Xia State Key Laboratory of Food Science and Resources School of Food Science and Technology Collaborative Innovation Center of Food Safety and Quality Control in Jiangsu Province Jiangnan University Lihu 1800 Wuxi Jiangsu 214122 China Department of Food Science and Technology National University of Singapore Science Drive 2 Singapore 117546 Singapore National University of Singapore (Suzhou) Research Institute No. 377 Linquan Street Suzhou Industrial Park Suzhou Jiangsu 215123 China

This study investigated the effect of 2-phenylethanol (2-PE), a quorum-sensing molecule, on the performance of Saccharomyces cerevisiae 31 in sour fish fermentation. Results demonstrated that 250 μM 2-PE significantly influenced yeast viability and metabolic activity. 2-PE supplementation altered the cell wall structure, increased membrane permeability, and promoted biofilm formation, which is crucial for enhancing fermentation flavor. 2-PE accelerated glucose consumption and induced a significant decrease in pH. Additionally, 2-PE supplementation led to a higher production of esters and alcohols, key contributors to the sour fish flavor profile. The OPLS-DA highlighted the distinct separation between the 2-PE-treated (250 μM) and untreated (0 μM) groups, suggesting a direct impact on the flavor composition. Thus, 2-PE supplementation can regulate the metabolic activity of S. cerevisiae 31, leading to a controlled and enhanced production of flavor compounds. This approach provides a novel strategy for improving the quality and consistency of traditional fermented fish.

关键词： 2-Phenylethanol Fermentation performance Quorum sensing Saccharomyces cerevisiae 31 Sour fish

来源：评论

学校读者我要写书评

暂无评论

Time-delay assisted mechanism and adaptive LSTM hybrid train braking model of heavy haul trains

引用

control Engineering Practice 2025年 163卷

作者： Qiang Liu Kexuan Xu Yating Fu Jiang Liu Ling Liu State Key Laboratory of Synthetical Automation for Process Industries (Northeastern University) Shenyang 110819 China Key Laboratory of Advanced Control and Optimization of Jiangxi Province Nanchang 330013 China School of Electrical and Automation Engineering East China Jiaotong University Nanchang 330013 China Frontiers Science Center for Smart High-speed Railway System Beijing 100044 China School of Electronic and Information Engineering Beijing Jiaotong University Beijing 100044 China Beijing National Railway Research & Design Institute of Signal Communication Group Company Beijing 100070 China

The train braking model (TBM) that describes the dynamic relations of operation speed, mileage, and control force is essential for achieving stable operation and precise stopping of heavy haul trains (HHTs). However, it is difficult to establish the TBM of HHTs due to complex characteristics: (i) the long body and air braking process of the HHTs may lead to unexpected time-delays of control force; and (ii) there are significant unmodeled dynamics caused by rough tracks and external poor environment. Traditional TBM does not take into account the unmodeled dynamics and time-delays caused by air transmission during braking. To address these issues, this study proposes a data mechanism hybrid modeling strategy, which incorporates a braking time-delay assisted mechanism model and an adaptive long and short-term memory (LSTM) model. A new Bayesian optimization based time-delay estimation method is first proposed to determine unknown time-delays of each carriage and the estimated time-delays are incorporated to generate the multi-point-mass kinetic mechanism model. Moreover, the error of the mechanism-driven model is adaptively compensated by a sliding window LSTM model to conduct the unmodeled dynamics. The effectiveness of the proposed method is demonstrated using the field data.

关键词： Railroad transportation

来源：评论

学校读者我要写书评

暂无评论

Facilitating renewable natural gas production for a circular bioeconomy: AI-Driven process visualization and data augmentation on biochar-mediated anaerobic digestion

引用

Chemical Engineering Journal 2025年 516卷

作者： Xiaoman He Jingyuan Guo Xihui Kang Xue Ning Huichao Chen Daolun Liang Chen Deng Zutan Li Dekui Shen Huiyan Zhang Richen Lin Jerry D. Murphy Key Laboratory of Energy Thermal Conversion and Control of Ministry of Education School of Energy and Environment Southeast University Nanjing 211189 China Advanced Biofuels and Bioproducts Process Development Unit Biological Systems and Engineering Division Lawrence Berkeley National Laboratory Emeryville CA 94608 USA School of Engineering and Architecture University College Cork T23XE10 Cork Ireland Research Ireland MaREI Centre for Energy Climate and Marine Environmental Research Institute University College Cork T23XE10 Cork Ireland School of Chemistry and Chemical Engineering Southeast University Nanjing 211189 China School of Engineering Westlake University Hangzhou Zhejiang 310030 China Institute of Advanced Technology Westlake Institute for Advanced Study Hangzhou Zhejiang 310024 China

The conversion of biomass residues into biochar is a promising strategy for enhancing sustainability within the circular bioeconomy, particularly through its role in improving renewable natural gas production. However, engineering biochar with optimal properties remains a complex challenge, as the relationship between preparation conditions, biochar characteristics, and anaerobic digestion (AD) performance is not fully understood. This study presents an AI-derived full-process prediction approach that integrates machine learning and generative models to guide the rational design of biochar, and optimize its use for biomethane production. Three tree-based regression models were employed to predict AD performance, with the eXtreme Gradient Boosting Regression model demonstrating superior accuracy. Feature importance analysis identified key biochar properties, including electrical conductivity, oxygen content, and specific surface area, as critical factors influencing biomethane production. These properties can be fine-tuned by adjusting pyrolysis conditions and selecting suitable biomass sources. A generative adversarial network was further used to explore a broader data space, helping to identify the optimal combination of parameters for maximizing AD efficiency. This novel AI-driven framework facilitates biochar-mediated renewable natural gas production, offering a scalable and sustainable approach for advancing circular bioeconomy.

关键词： Biochar-mediated anaerobic digestion Circular bioeconomy Data augmentation Machine learning Renewable natural gas

来源：评论

学校读者我要写书评

暂无评论

Synergistic biotechnological co-processing of wastewater and external sludge for partial energy autarky

引用

Chemical Engineering Journal 2025年

作者： Fangzhai Zhang Hongying Lu Shouyou Gao Dan Qu Yujing Zhang Ziyi Du Jiahui Wang Yongzhen Peng National Engineering Laboratory for Advanced Municipal Wastewater Treatment and Reuse Technology Engineering Research Center of Beijing Beijing University of Technology Beijing 100124 China Beijing General Municipal Engineering Design & Research Institute Beijing 100082 China Beijing Key Laboratory for Source Control Technology of Water Pollution College of Environmental Science and Engineering Beijing Forestry University 35 Qinghua East Road Haidian District Beijing 100083 China Engineering Research Center for Water Pollution Source Control & Eco-remediation College of Environmental Science and Engineering Beijing Forestry University 35 Qinghua East Road Haidian District Beijing 100083 China

Harnessing bioenergy from waste activated sludge (WAS) is crucial for sustainable development in wastewater treatment field; however, traditional anaerobic digestion with pretreatment is energy-intensive. We developed a novel, partially energy-autonomous biotechnology that synergistically co-processes wastewater and WAS using a self-generated biocatalyst (ppm-level FNA) without any external energy or chemical input. This self-sustained FNA-centric system simultaneously enables rapid organic resource recovery, sludge disintegration, and denitrification within only 20.5 h, representing a major advancement over traditional multi-day anaerobic processes. The multifunctional process recovered 3264.7 mg/L of net organics, reduced sludge by 7.92 kg/m 3 ·d, and removed nitrogen at a rate of 0.71 kg N/m 3 ·d with 92.5 % efficiency. A 58.4 % decrease in cell membrane thickness (from 10.42 nm to 4.7 nm) and significant perforation were observed under biocatalyst treatment. Comprehensive analysis, including cell surface morphology, microorganism survival status, organic resource dynamic, and proteins spatial distribution, underscore organic resource recovery from sludge relying on multifaceted biocatalytic and stress effects. Consequently, even after biocatalyst completed consumption, organic release rate remained constant at 87.7–122 mg COD/h. Furthermore, unlike the detrimental effects on biodiversity, specific microbial communities were selectively enriched, remarkable enhancing in glycolysis, fatty acid synthesis, and denitrification functionalities. Results demonstrate feasibility of partial energy self-sufficiency, contributing to achievement of carbon neutrality in wastewater and WAS management.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Batch Normalization Masked Sparse Autoencoder for Robotic Grasping Detection

Batch Normalization Masked Sparse Autoencoder for Robotic Gr...

引用

2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

作者： Zhenzhou Shao Ying Qu Guangli Ren Guohui Wang Yong Guan Zhiping Shi Jindong Tan Beijing Advanced Innovation Center for Imaging Technology and Beijing Key Laboratory of Light Industrial Robot and Safety Verification Capital Normal University Beijing China The Engineering College The University of Tennessee Knoxville TN USA State Key Laboratory for Management and Control of Complex Systems Institute of Automation Beijing China

ISBN: (数字)9781728162126

ISBN: (纸本)9781728162133

To improve the accuracy of the grasping detection, this paper proposes a novel detector with batch normalization masked evaluation model. It is designed with a two-layer sparse autoencoder, and a Batch Normalization based mask is incorporated into the second layer of the model to effectively reduce the features with weak correlation. The extracted features from such model are more distinctive, which guarantees the higher accuracy of the grasping detection. Extensive experiments show that the proposed evaluation model outperforms the state-of- the-art, and the recognition accuracy can reach 95.51% for robotic grasping detection.

关键词： Training Redundancy Grasping Detectors Feature extraction Intelligent robots Load modeling

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：