检索结果-内蒙古大学图书馆

ASDFL: An adaptive super-pixel discriminative feature-selective learning for vehicle matching

EXPERT SYSTEMS 2023年第2期40卷 e13144-e13144页

作者： Qin, Rong Lv, Huanhuan Zhang, Yi huangfu, luwen huang, Sheng Chongqing Univ Sch Big Data & Software Engn Chongqing Peoples R China Chongqing Univ Key Lab Dependable Serv Comp Cyber Phys Soc Minist Educ Chongqing Peoples R China San Diego State Univ Fowler Coll Business San Diego CA 92182 USA San Diego State Univ Ctr Human Dynam Mobile Age San Diego CA 92182 USA

There are a large number of cameras in modern transportation system that capture numerous vehicle images continuously. Therefore, automatic analysis of these vehicle images is helpful for traffic flow management, criminal investigations and vehicle inspections. Vehicle matching, which aims to determine whether two input images depict an identical vehicle, is one of the core tasks in vehicle analysis. Recent relevant studies have focused on local feature extraction instead of global extraction, since local details can provide crucial cues to distinguish between cars. However, these methods do not select local features;that is, they do not assign weights to local features. Therefore, in this research, we systematically study the vehicle matching task, and present a novel annotation-free local-based deep learning method called Adaptive super-pixel discriminative feature-selective learning (ASDFL) to address this issue. In ASDFL, vehicle images are segmented into clusters of super-pixels of similar size by considering the location and colour similarities of pixels without using any component-level annotation. These super-pixels are deemed to be the virtual components of vehicles. Moreover, a convolutional neural network is used to extract the deep features of these virtual components. Thereafter, an instance-specific mask generation module driven by the extracted global features is enhanced to produce a mask to select the most distinctive virtual components of each vehicle image pair in the feature space. Finally, the vehicle matching task is accomplished by classifying the selected virtual component features of each imaged vehicle pair. Extensive experiments on two popular vehicle identification benchmarks demonstrate that our method is 1.57% and 0.8% more accurate than the previous baselines in a vehicle matching task on the VeRi and VehicleID datasets, respectively, which demonstrates the effectiveness of our method.

关键词： feature-selective learning intelligent transportation systems super-pixel segmentation vehicle matching vehicle re-identification

来源：评论

学校读者我要写书评

暂无评论

Mitochondria: fundamental characteristics, challenges, and impact on aging

引用

BIOGERONTOLOGY 2024年第6期25卷 923-941页

作者： Liang, Runyu Zhu, luwen huang, Yongyin Chen, Jia Tang, Qiang Heilongjiang Univ Chinese Med Harbin Peoples R China Heilongjiang Univ Chinese Affiliated Hosp 2 Harbin Peoples R China

As one of the most vital organelles within biological cells, mitochondria hold an irreplaceable status and play crucial roles in various diseases. Research and therapies targeting mitochondria have achieved significant progress in numerous conditions. Throughout an organism's lifespan, mitochondrial dynamics persist continuously, and due to their inherent characteristics and various external factors, mitochondria are highly susceptible to damage. This susceptibility is particularly evident during aging, where the decline in biological function is closely intertwined with mitochondrial dysfunction. Despite being an ancient and enigmatic organelle, much remains unknown about mitochondria. Here, we will explore the past and present knowledge of mitochondria, providing a comprehensive review of their intrinsic properties and interactions with nuclear DNA, as well as the challenges and impacts they face during the aging process.

关键词： Mitochondria Aging Endosymbiosis Mitochondria DNA Nuclear DNA

来源：评论

学校读者我要写书评

暂无评论

脑脊液二代测序技术确诊单核细胞增生李斯特菌脑膜炎两例报道

引用

实用心脑肺血管病杂志 2022年第10期30卷 133-136页

作者：陈平苟惠吴文兵黄露雯任思冲四川省遂宁市中心医院药学部 629000 西南医科大学附属医院药学部四川省泸州市646000 西南医科大学基础医学院生物化学与分子生物学教研室四川省泸州市646000 四川省遂宁市中心医院神经内科 629000 成都医学院第一附属医院肾内科四川省成都市610500

单核细胞增生李斯特菌脑膜炎临床少见,病死率高,传统检测方法不易获得其病原学证据,且单核细胞增生李斯特菌对头孢菌素类抗生素天然耐药,因而其早期诊断与精准治疗仍是临床面临的挑战。二代测序技术使得单核细胞增生李斯特菌脑膜炎病原... 详细信息

单核细胞增生李斯特菌脑膜炎临床少见,病死率高,传统检测方法不易获得其病原学证据,且单核细胞增生李斯特菌对头孢菌素类抗生素天然耐药,因而其早期诊断与精准治疗仍是临床面临的挑战。二代测序技术使得单核细胞增生李斯特菌脑膜炎病原菌检出率不断提高,本文报道2例通过脑脊液二代测序成功确诊单核细胞增生李斯特菌脑膜炎病例,并及时给予合理抗菌药物治疗,患者预后良好,以期为临床提供参考。

关键词：脑膜炎,利斯特菌性单核细胞增生李斯特菌脑脊液二代测序病例报告

来源：评论

学校读者我要写书评

暂无评论

Data Distribution Distilled Generative Model for Generalized Zero-Shot Recognition 38

Data Distribution Distilled Generative Model for Generalized...

引用

38th AAAI Conference on Artificial Intelligence (AAAI) / 36th Conference on Innovative Applications of Artificial Intelligence / 14th Symposium on Educational Advances in Artificial Intelligence

作者： Wang, Yijie Hong, Mingjian huangfu, luwen huang, Sheng Chongqing Univ Chongqing Peoples R China San Diego State Univ Fowler Coll Business San Diego CA 92182 USA

ISBN: (纸本)1577358872

In the realm of Zero-Shot Learning (ZSL), we address biases in Generalized Zero-Shot Learning (GZSL) models, which favor seen data. To counter this, we introduce an end-to-end generative GZSL framework called D(3)GZSL. This framework respects seen and synthesized unseen data as in-distribution and out-of-distribution data, respectively, for a more balanced model. D(3)GZSL comprises two core modules: in-distribution dual space distillation ((IDSD)-S-2) and out-of-distribution batch distillation ((ODBD)-D-2). (IDSD)-S-2 aligns teacher-student outcomes in embedding and label spaces, enhancing learning coherence. (ODBD)-D-2 introduces low-dimensional out-of-distribution representations per batch sample, capturing shared structures between seen and unseen categories. Our approach demonstrates its effectiveness across established GZSL benchmarks, seamlessly integrating into mainstream generative frameworks. Extensive experiments consistently showcase that D(3)GZSL elevates the performance of existing generative GZSL methods, underscoring its potential to refine zero-shot learning *** code is available at: https://***/PJBQ/***

关键词： Zero-shot learning

来源：评论

学校读者我要写书评

暂无评论

See My Sea - Emotion Translation and Visualization 24

See My Sea - Emotion Translation and Visualization

引用

17th International Symposium on Visual Information Communication and Interaction

作者： Jiang, Yuyang huang, Zhuoxin Yu, luwen Li, Hao Hui, Pan Hong Kong Univ Sci & Technol Guangzhou Guangzhou Peoples R China Hong Kong Univ Sci & Technol Hong Kong Peoples R China Lingnan Univ Hong Kong Peoples R China

ISBN: (纸本)9798400709678

"See My Sea" is an innovative project that integrates art and emotion to explore human emotions through facial expressions. Utilizing advanced facial recognition and real-time data processing, the project captures subtle emotional cues and translates them into dynamic visualizations of ocean particles. This artistic endeavor transforms transient facial movements into fluid, evolving visual patterns, mirroring the vast and ever-changing nature of the sea. Viewers engage interactively with their emotional states, influencing the visual output and fostering a deeper understanding of their inner selves.

关键词： Somatosensory Interaction Facial Emotion Emotion Design Mental Health

来源：评论

学校读者我要写书评

暂无评论

Clinical outcomes of modified transarterial chemoembolization vs. drug-eluting bead transarterial chemoembolization as initial treatment for single hepatocellular carcinoma: a propensity score matching analysis

引用

ANNALS OF MEDICINE 2024年第1期56卷 2423787页

作者： Chang, Boyang Zhou, Churen Liu, Chengcheng Mu, luwen huang, Mingsheng Bai, Mingjun Sun Yat Sen Univ Affiliated Hosp 3 Dept Intervent Radiol Guangzhou 510630 Guangdong Peoples R China Sun Yat Sen Univ Affiliated Hosp 3 Dept Hematol Guangzhou Guangdong Peoples R China

Background & aims: Optimizing transarterial chemoembolization (TACE) can enhance treatment efficacy for hepatocellular carcinoma (HCC). This study compares modified TACE (M-TACE), which combines a lipiodol-based emulsion and drug-eluting beads, with drug-eluting bead TACE (DEB-TACE) as initial therapies for solitary HCC. Methods: In this retrospective study, 185 patients undergoing M-TACE or DEB-TACE were evaluated. Propensity score matching was used to create 69 balanced pairs. Initial tumor response, repeated treatments within six months, local tumor progression (LTP), overall survival (OS), and adverse events (AEs) were assessed. Results: M-TACE exhibited significantly higher initial complete response (CR) rates (39.1% vs. 23.2%, p = 0.043) and fewer repeated treatments within six months (1.7 +/- 0.9 vs. 2.1 +/- 0.7;p = 0.033) compared to DEB-TACE. LTP rates were notably lower with M-TACE at 12 months (39.1% vs. 65.2%, p = 0.002), and median time to LTP was prolonged with M-TACE (13.3 vs. 8.2 months, p = 0.038). Stratified analysis revealed a significantly longer OS in individuals achieving a CR after the initial M-TACE (50.5 vs. 33.4 months, p = 0.043). However, the overall study population did not exhibit a significant difference in OS between the two groups. Comparable AEs (all p > 0.05) were observed. Conclusions: M-TACE showed higher initial CR rates, lower LTP rates, and extended time to LTP compared to DEB-TACE, indicating its potential to enhance TACE effectiveness for solitary HCC.

关键词： Carcinoma hepatocellular treatment outcome chemoembolization therapeutic

来源：评论

学校读者我要写书评

暂无评论

Conversion of carbon dioxide to carbon monoxide: Two-step chemical looping dry reforming using Ca2Fe2O5-Zr0.5Ce0.5O2 composite oxygen carriers

引用

FUEL 2022年 322卷 1页

作者： Wang, luwen Lin, Yan huang, Zhen Zeng, Kuo huang, Hongyu Huazhong Univ Sci & Technol Sch Energy & Power Engn State Key Lab Coal Combust Wuhan 430074 Peoples R China Chinese Acad Sci CAS Guangzhou Inst Energy Convers Key Lab Renewable Energy 2 Nengyuan Rd Guangzhou 510640 Peoples R China Southern Marine Sci & Engn Guangdong Lab Guangzho 1119 Haibin Rd Guangzhou 511458 Peoples R China

In this study, Zr0.5Ce0.5O2 (ZC) was used to improve the reactivity of Ca2Fe2O5 (CF) oxygen carriers (OCs) and alleviate sintering. The reactivity of modified OCs in Chemical looping dry reforming (CLDR) was explored using a thermogravimetric analyzer and a fixed-bed reactor, and the fresh and reacted OCs were characterized by X-ray diffraction (XRD) and transmission electron microscopy (TEM). The results revealed that CF6 (60 wt% CF + 40 wt% ZC) had the best reactivity. Owing to the addition of ZC, the carbon monoxide (CO) yield of CF6 was twice that of CF during carbon dioxide (CO2) oxidation, increasing from 72.3 mL/g to 138 mL/g. In the cyclic experiments, the CO yield from CO2 oxidation decreased from 138 mL/g in the first cycle to 34.3 mL/g in the fifth cycle and subsequently stabilized over the next 15 cycles. The decrease in cyclic performance is related to the sintering of the sample and the formation of a new phase, CaZrO3, which hindered the redox reaction in the sample. Although CF6 had a decreased cyclic performance, it still had a better reactivity than CF, which had a CO yield that was about 20% lower than that of CF6. The reason for the improvement is that ZC improved the dispersibility of CF. After 20 cycles, the generation of CO was facilitated by extending the reduction time of the OC, and the CO yield increased from 35 mL/g to 60 mL/g when the reduction time was doubled.

关键词： Chemical looping dry reforming (CLDR) Oxygen carrier CO2 splitting Ca2Fe2O5 Zr0.5Ce0.5O2

来源：评论

学校读者我要写书评

暂无评论

Effect of ultrasonic surface rolling on the surface morphology and hardness of dual-phase LZ91 alloy

引用

MATERIALS SCIENCE AND TECHNOLOGY 2025年

作者： huang, Zeyu Hu, luwen Fang, Xiaoying Cao, Zhi Yang, Huimin Wei, Guobing Xue, Zhang Chongqing Univ Coll Mat Sci & Engn Chongqing 400044 Peoples R China Chongqing Univ Natl Engn Res Ctr Magnesium Alloys Chongqing Peoples R China Chongqing Univ Int Joint Lab Light Alloys MOE Chongqing Peoples R China Chongqing Univ Natl Key Lab Adv Casting Technol Chongqing Peoples R China

Dual-phase Mg-Li alloys are promising for industrial applications due to their low density and superior plasticity, but their low hardness limits performance under high-stress conditions. This study employed ultrasonic surface rolling processing (USRP) to enhance the surface properties of the dual-phase LZ91 alloy. The effects of USRP on surface morphology, microstructure, and hardness were systematically investigated. Results showed that USRP generated a similar to 400 mu m gradient surface layer. TEM revealed grain refinement to similar to 300 nm in both beta-Li and alpha-Mg phases. Surface hardness increased from 52.7 HV to 82.6 HV (56.7% improvement) under identical loading. These findings demonstrate that USRP effectively strengthens the LZ91 alloy by refining grains and forming gradient structures.

关键词： ultrasonic surface rolling process dual-phase Mg-Li alloy microstructure surface hardness gradient structure

来源：评论

学校读者我要写书评

暂无评论

Ascending aorta dilatation indicates obstructive sleep apnea in Chinese patients with cardiovascular disease

引用

SLEEP AND BREATHING 2025年第1期29卷 1-7页

作者： Qiu, luwen Li, Pingwei Ding, Zhaohui Wang, Jing huang, Ning Hua, Lu Xi, Qunying Chinese Acad Med Sci Pulm Vasc Dis Ward Fuwai Hosp 12 Langshan Rd Shenzhen 518057 Guangdong Provi Peoples R China Chinese Acad Med Sci Ward 2 Cardiac Arrhythmia Fuwai Hosp Shenzhen 518057 Guangdong Provi Peoples R China

PurposeObstructive sleep apnea (OSA) is prevalent in patients with cardiovascular disease (CVD), and is linked to worsened outcomes. Screening tools are essential for early detection and intervention. We aimed to investigate the role of ascending aorta diameter (AAD) obtained from echocardiography in identifying OSA in CVD patients. MethodsWe examined the correlation between AAD and sleep apnea test parameters in 721 hospitalized CVD (including hypertension, coronary artery disease, chronic heart failure, atrial fibrillation, and pulmonary hypertension) patients. Multivariate analyses were performed on all findings. Receiver operator characteristic curve (ROC) analyses were used to determine the role of AAD in detecting OSA. ResultsAAD was significantly higher in patients with OSA than those without OSA. AAD significantly correlated with apnea-hypopnea index (AHI) and saturation-related parameters. In comparison, pulmonary artery diameter did not correlate with AHI but had significant negative correlations with the saturation-related parameters. The area under the ROC of AAD for detecting OSA in CVD patients was 0.70 (95% CI, 0.65-0.74;p < 0.001). In the chronic heart failure subgroup, the area under the curve for AAD was 0.75 (95% CI 0.67-0.82, p < 0.001). The optimal cut-off value of AAD was 31.5 mm, resulting in a sensitivity of 73%, a specificity of 55% in CVD patients, and 29.5 mm with a sensitivity of 82% and a specificity of 64% in chronic heart failure patients. ConclusionsAAD was associated with the severity of OSA. AAD could be a valuable tool for identifying OSA in CVD patients, particularly in patients with chronic heart failure.

关键词： Obstructive sleep apnea Echocardiography Ascending aorta diameter Diagnosis

来源：评论

学校读者我要写书评

暂无评论

SAM-MIL: A Spatial Contextual Aware Multiple Instance Learning Approach for Whole Slide Image Classification 24

SAM-MIL: A Spatial Contextual Aware Multiple Instance Learni...

引用

32nd ACM International Conference on Multimedia, MM 2024

作者： Fang, Heng huang, Sheng Tang, Wenhao huangfu, luwen Liu, Bo School of Big Data & Software Engineering Chongqing University Chongqing China Fowler College of Business San Diego State University San DiegoCA United States School of Computer Science and Information Engineering Hefei University of Technology Hefei China

ISBN: (纸本)9798400706868

Multiple Instance Learning (MIL) represents the predominant framework in Whole Slide Image (WSI) classification, covering aspects such as sub-typing, diagnosis, and beyond. Current MIL models predominantly rely on instance-level features derived from pretrained models such as ResNet. These models segment each WSI into independent patches and extract features from these local patches, leading to a significant loss of global spatial context and restricting the model's focus to merely local features. To address this issue, we propose a novel MIL framework, named SAM-MIL, that emphasizes spatial contextual awareness and explicitly incorporates spatial context by extracting comprehensive, image-level information. The Segment Anything Model (SAM) represents a pioneering visual segmentation foundational model that can capture segmentation features without the need for additional fine-tuning, rendering it an outstanding tool for extracting spatial context directly from raw WSIs. Our approach includes the design of group feature extraction based on spatial context and a SAM-Guided Group Masking strategy to mitigate class imbalance issues. We implement a dynamic mask ratio for different segmentation categories and supplement these with representative group features of categories. Moreover, SAM-MIL divides instances to generate additional pseudo-bags, thereby augmenting the training set, and introduces consistency of spatial context across pseudo-bags to further enhance the model's performance. Experimental results on the CAMELYON-16 and TCGA Lung Cancer datasets demonstrate that our proposed SAM-MIL model outperforms existing mainstream methods in WSIs classification. Our open-source implementation code is is available at https://***/FangHeng/SAM-MIL. © 2024 ACM.

关键词： Self-supervised learning

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：