检索结果-内蒙古大学图书馆

arXiv 2019年

作者： Zhang, Zhuosheng Wu, Yuwei Zhou, Junru Duan, Sufeng Zhao, Hai Wang, Rui Department of Computer Science and Engineering Shanghai Jiao Tong University Shanghai China Key Lab. of Shanghai Education Commission for Intelligent Interaction and Cognitive Engineering Shanghai Jiao Tong University Shanghai China MoE Key Lab of Artificial Intelligence AI Institute Shanghai Jiao Tong University Shanghai China College of Zhiyuan Shanghai Jiao Tong University China Kyoto Japan

For machine reading comprehension, the capacity of effectively modeling the linguistic knowledge from the detail-riddled and lengthy passages and getting ride of the noises is essential to improve its performance. Traditional attentive models attend to all words without explicit constraint, which results in inaccurate concentration on some dispensable words. In this work, we propose using syntax to guide the text modeling of both passages and questions by incorporating explicit syntactic constraints into attention mechanism for better linguistically motivated word representations. To serve such a purpose, we propose a novel dual contextual architecture called syntax-guided network (SG-Net), which consists of a BERT context vector and a syntax-guided context vector, to provide more fine-grained representation. Extensive experiments on popular benchmarks including SQuAD 2.0 and RACE show that the proposed approach achieves a substantial and significant improvement over the fine-tuned BERT baseline. Copyright © 2019, The Authors. All rights reserved.

关键词： Syntactics

来源：评论

学校读者我要写书评

暂无评论

Correction to: Editorial

引用

International Journal of Social Robotics 2020年第1期12卷 297-297页

作者： Agnieszka Wykowska Shuzhi Sam Ge Oussama Khatib Social Cognition in Human-Robot Interaction Istituto Italiano Di Tecnologia Genoa Italy Social Robotics Laboratory Interactive Digital Media Institute The National University of Singapore Singapore Singapore Centre for Robotics UESTC Chengdu China Robotics Laboratory Department of Computer Science Stanford University Stanford USA

The following mistakes were inadvertently overlooked in the production of the Editorial of the January issue of Social Robotics.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Designing ecosystems of intelligence from first principles

引用

Collective Intelligence 2024年第1期3卷

作者： Karl J Friston Maxwell JD Ramstead Alex B Kiefer Alexander Tschantz Christopher L Buckley Mahault Albarracin Riddhi J Pitliya Conor Heins Brennan Klein Beren Millidge Dalton AR Sakthivadivel Toby St Clere Smithe Magnus Koudahl Safae Essafi Tremblay Capm Petersen Kaiser Fung Jason G Fox Steven Swanson Dan Mapes Gabriel René VERSES Research Lab Los Angeles CA USA (KJF MJDR ABK AT CLB MA RJP CH BK BM DARS TSCS MK SET CP KF JGF SS DM GR) Wellcome Centre for Human Neuroimaging University College London London UK (KJF MJDR) Department of Philosophy Monash University Melbourne VIC Australia (ABK) Department of Informatics University of Sussex Brighton UK (CLB) Department of Computer Science Université du Québec à Montréal Montréal QC Canada (MA) Department of Experimental Psychology University of Oxford Oxford UK (RJP TSCS) Department of Collective Behaviour Max Planck Institute of Animal Behavior Konstanz Germany (CH) Department of Biology University of Konstanz Konstanz Germany (CH) Centre for the Advanced Study of Collective Behaviour University of Konstanz Konstanz Germany (CH) Network Science Institute Northeastern University Boston MA USA (BK) Brain Network Dynamics Unit University of Oxford Oxford UK (BM) Department of Mathematics Stony Brook University New York NY USA (DARS) Department of Physics and Astronomy Stony Brook University New York NY USA (DARS) Department of Biomedical Engineering Stony Brook University New York NY USA (DARS) Topos Institute Berkeley CA USA (TSCS) Department of Electrical Engineering Eindhoven University of Technology Eindhoven Netherlands (MK) Department of Philosophy Université du Québec à Montréal Montréal QC Canada (SET)

This white paper lays out a vision of research and development in the field of artificial intelligence for the next decade (and beyond). Its denouement is a cyber-physical ecosystem of natural and synthetic sense-making, in which humans are integral participants—what we call “shared intelligence.” This vision is premised on active inference, a formulation of adaptive behavior that can be read as a physics of intelligence, and which inherits from the physics of self-organization. In this context, we understand intelligence as the capacity to accumulate evidence for a generative model of one’s sensed world—also known as self-evidencing. Formally, this corresponds to maximizing (Bayesian) model evidence, via belief updating over several scales, that is, inference, learning, and model selection. Operationally, this self-evidencing can be realized via (variational) message passing or belief propagation on a factor graph. Crucially, active inference foregrounds an existential imperative of intelligent systems; namely, curiosity or the resolution of uncertainty. This same imperative underwrites belief sharing in ensembles of agents, in which certain aspects (i.e., factors) of each agent’s generative world model provide a common ground or frame of reference. Active inference plays a foundational role in this ecology of belief sharing—leading to a formal account of collective intelligence that rests on shared narratives and goals. We also consider the kinds of communication protocols that must be developed to enable such an ecosystem of intelligences and motivate the development of a shared hyper-spatial modeling language and transaction protocol, as a first—and key—step towards such an ecology.

关键词： Active inference free energy principle artificial intelligence belief updating belief propagation

来源：评论

学校读者我要写书评

暂无评论

On modular training of neural acoustics-to-word model for LVCSR

arXiv

引用

arXiv 2018年

作者： Chen, Zhehuai Liu, Qi Li, Hao Yu, Kai Key Lab. of Shanghai Education Commission for Intelligent Interaction and Cognitive Engineering SpeechLab Department of Computer Science and Engineering Shanghai Jiao Tong University China

End-to-end (E2E) automatic speech recognition (ASR) systems directly map acoustics to words using a unified model. Previous works mostly focus on E2E training a single model which integrates acoustic and language model into a whole. Although E2E training benefits from sequence modeling and simplified decoding pipelines, large amount of transcribed acoustic data is usually required, and traditional acoustic and language modelling techniques cannot be utilized. In this paper, a novel modular training framework of E2E ASR is proposed to separately train neural acoustic and language models during training stage, while still performing end-to-end inference in decoding stage. Here, an acoustics-to-phoneme model (A2P) and a phoneme-to-word model (P2W) are trained using acoustic data and text data respectively. A phone synchronous decoding (PSD) module is inserted between A2P and P2W to reduce sequence lengths without precision loss. Finally, modules are integrated into an acoustics-to-word model (A2W) and jointly optimized using acoustic data to retain the advantage of sequence modeling. Experiments on a 300-hour Switchboard task show significant improvement over the direct A2W model. The efficiency in both training and decoding also benefits from the proposed *** Codes 68T10 Copyright © 2018, The Authors. All rights reserved.

关键词： Decoding

来源：评论

学校读者我要写书评

暂无评论

Second workshop on theory-informed user modeling for tailoring and personalizing interfaces (humanIZE): Workshop preface

Second workshop on theory-informed user modeling for tailori...

引用

2018 Joint ACM IUI Workshops, ACMIUI-WS 2018

作者： Graus, Mark Ferwerda, Bruce Tkalcic, Marko Germanakos, Panagiotis Human-Technology Interaction Group Eindhoven University of Technology P.O. Box 513 Eindhoven NL5600 MB Netherlands Department of Computer Science and Informatics School of Engineering Jönköping University P.O. Box 1026 Jönköping SESE-551 11 Sweden Faculty of Computer Science Free University of Bozen-Bolzano Piazza Domenicani 3 Bozen-Bolzano ITI-39100 Italy UX Mobile and Business Services PandI ICD SAP SE Dietmar-Hopp-Allee 16 Walldorf DE69190 Germany Department of Computer Science University of Cyprus P.O. Box 20537 Nicosia1678 Cyprus

The second workshop on Theory-Informed UserModeling for Tailoring and Personalizing Interfaces (humanIZE)1 took place in conjunction with the 23rd annual meeting of the intelligent user interfaces (IUI)2 community in Tokyo, Japan on March 11, 2018. The goal of the workshop was to attract researchers from different fields by accepting contributions on the intersection of practical data mining methods and theoretical knowledge for personalization. A total of eight papers were accepted for this edition of the workshop. © 2018 Copyright for the individual papers remains with the authors.

关键词： User interfaces

来源：评论

学校读者我要写书评

暂无评论

ON MODULAR TRAINING OF NEURAL ACOUSTICS-TO-WORD MODEL FOR LVCSR

ON MODULAR TRAINING OF NEURAL ACOUSTICS-TO-WORD MODEL FOR LV...

引用

IEEE International Conference on Acoustics, Speech and Signal Processing

作者： Zhehuai Chen Qi Liu Hao Li Kai Yu Key Lab. of Shanghai Education Commission for Intelligent Interaction and Cognitive Engineering SpeechLab Department of Computer Science and Engineering Shanghai Jiao Tong University China

ISBN: (纸本)9781538646595

End-to-end (E2E) automatic speech recognition (ASR) systems directly map acoustics to words using a unified model. Previous works mostly focus on E2E training a single model which integrates a-coustic and language model into a whole. Although E2E training benefits from sequence modeling and simplified decoding pipelines, large amount of transcribed acoustic data is usually required, and traditional acoustic and language modelling techniques cannot be utilized. In this paper, a novel modular training framework of E2E ASR is proposed to separately train neural acoustic and language models during training stage, while still performing end-to-end inference in decoding stage. Here, an acoustics-to-phoneme model (A2P) and a phoneme-to-word model (P2W) are trained using acoustic data and text data respectively. A phone synchronous decoding (PSD) module is inserted between A2P and P2W to reduce sequence lengths without precision loss. Finally, modules are integrated into an acoustics-to-word model (A2W) and jointly optimized using acoustic data to retain the advantage of sequence modeling. Experiments on a 300-hour Switchboard task show significant improvement over the direct A2W model. The efficiency in both training and decoding also benefits from the proposed method.

关键词： E2E Connectionist Temporal Classification Attention-based Encoder-decoder Phone Synchronous Decoding end-to-end Acoustic data modelling languages DECODING Acoustics Speech recognition Training

来源：评论

学校读者我要写书评

暂无评论

Semantics-aware BERT for language understanding

arXiv

引用

arXiv 2019年

作者： Zhang, Zhuosheng Wu, Yuwei Zhao, Hai Li, Zuchao Zhang, Shuailiang Zhou, Xi Zhou, Xiang Department of Computer Science and Engineering Shanghai Jiao Tong University Key Laboratory of Shanghai Education Commission for Intelligent Interaction and Cognitive Engineering Shanghai Jiao Tong University Shanghai China MoE Key Lab of Artificial Intelligence AI Institute Shanghai Jiao Tong University Shanghai China College of Zhiyuan Shanghai Jiao Tong University China CloudWalk Technology Shanghai China

The latest work on language representations carefully integrates contextualized features into language model training, which enables a series of success especially in various machine reading comprehension and natural language inference tasks. However, the existing language representation models including ELMo, GPT and BERT only exploit plain context-sensitive features such as character or word embeddings. They rarely consider incorporating structured semantic information which can provide rich semantics for language representation. To promote natural language understanding, we propose to incorporate explicit contextual semantics from pre-trained semantic role labeling, and introduce an improved language representation model, Semantics-aware BERT (SemBERT), which is capable of explicitly absorbing contextual semantics over a BERT backbone. SemBERT keeps the convenient usability of its BERT precursor in a light fine-tuning way without substantial task-specific modifications. Compared with BERT, semantics-aware BERT is as simple in concept but more powerful. It obtains new state-of-the-art or substantially improves results on ten reading comprehension and language inference tasks. Copyright © 2019, The Authors. All rights reserved.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

The evolutionary history of 2,658 cancers (vol 578, pg 122, 2020)

引用

NATURE 2023年第7948期614卷 E42-E42页

作者： Gerstung, Moritz Jolly, Clemency Leshchiner, Ignaty Dentro, Stefan C. Gonzalez, Santiago Rosebrock, Daniel Mitchell, Thomas J. Rubanova, Yulia Anur, Pavana Yu, Kaixian Tarabichi, Maxime Deshwar, Amit Wintersinger, Jeff Kleinheinz, Kortine Vazquez-Garcia, Ignacio Haase, Kerstin Jerman, Lara Sengupta, Subhajit Macintyre, Geoff Malikic, Salem Donmez, Nilgun Livitz, Dimitri G. Cmero, Marek Demeulemeester, Jonas Schumacher, Steven Fan, Yu Yao, Xiaotong Lee, Juhee Schlesner, Matthias Boutros, Paul C. Bowtell, David D. Zhu, Hongtu Getz, Gad Imielinski, Marcin Beroukhim, Rameen Sahinalp, S. Cenk Ji, Yuan Peifer, Martin Markowetz, Florian Mustonen, Ville Yuan, Ke Wang, Wenyi Morris, Quaid D. Spellman, Paul T. Wedge, David C. Van Loo, Peter European Molecular Biology Laboratory European Bioinformatics Institute (EMBL-EBI) Cambridge UK European Molecular Biology Laboratory Genome Biology Unit Heidelberg Germany Wellcome Sanger Institute Cambridge UK Genome Biology Unit European Molecular Biology Laboratory (EMBL) Heidelberg Germany University of Ljubljana Ljubljana Slovenia The Francis Crick Institute London UK Big Data Institute University of Oxford Oxford UK Wellcome Sanger Institute Wellcome Genome Campus Hinxton UK Big Data Institute Li Ka Shing Centre University of Oxford Oxford UK University of Cambridge Cambridge UK Cambridge University Hospitals NHS Foundation Trust Cambridge UK Department of Applied Mathematics and Theoretical Physics Centre for Mathematical Sciences University of Cambridge Cambridge UK Department of Epidemiology and Biostatistics Memorial Sloan Kettering Cancer Center New York NY USA Department of Statistics Columbia University New York NY USA Department of Haematology University of Cambridge Cambridge UK University of Leuven Leuven Belgium Broad Institute of MIT and Harvard Cambridge MA USA Center for Cancer Research Massachusetts General Hospital Charlestown MA USA Department of Pathology Massachusetts General Hospital Boston MA USA Harvard Medical School Boston MA USA Center for Cancer Research Massachusetts General Hospital Boston MA USA Dana-Farber Cancer Institute Boston MA USA Department of Medical Oncology Dana-Farber Cancer Institute Boston MA USA Department of Cancer Biology Dana-Farber Cancer Institute Boston MA USA Oxford NIHR Biomedical Research Centre Oxford UK Oxford NIHR Biomedical Research Centre University of Oxford Oxford UK University of Toronto Toronto Ontario Canada Vector Institute Toronto Ontario Canada Vector Institute Toronto ON Canada Department of Computer Science University of Toronto Toronto ON Canada Ontario Institute for Cancer Research Toronto Ontario Canada University of California Los Angeles

来源：评论

学校读者我要写书评

暂无评论

Association of BRAF V600E with Hypothalamic-Pituitary-Adrenal Axis Involvement in Erdheim-Chester Disease

引用

Endocrine and Metabolic science 2020年第1-2期1卷

作者： Hannah-Shmouni, Fady Boyd, Louisa Papadakis, Georgios Z. Tirosh, Amit Gahl, William A Estrada-Veras, Juvianee I. Section on Endocrinology and Genetics Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) National Institutes of Health Bethesda MD United States Office of the Clinical Director National Human Genome Research Institute National Institutes of Health Bethesda MD United States Department of Medical Imaging Heraklion University Hospital Medical School University of Crete Heraklion 70013 Greece Computational Biomedicine Laboratory (CBML) Institute of Computer Science (ICS) Foundation for Research and Technology Hellas (FORTH) Heraklion 70013 Greece Endocrine Oncology Bioinformatics Lab Department of Endocrinology Sheba Medical Center and Sackler Faculy of Medicine Tel Aviv University Tel Hashomer Israel Section on Human Biochemical Genetics Medical Genetics Branch National Human Genome Research Institute National Institutes of Health Bethesda MD United States

Erdheim-Chester disease (ECD) is a rare non-Langerhans cell histiocytosis characterized by uncontrolled inflammation resulting in fibro-inflammatory damage of multiple organs, including the hypothalamic-pituitary-adrenal (HPA) axis. Glucocorticoid insufficiency may occur and is associated with an increased burden of disease and an increased risk of adrenal crisis. Sixty-one ECD subjects (46 males, age 54.3 ± 10.8 years) were evaluated;56.1% (32/57) harbored a somatic activating BRAF V600E variant. Adrenal gland and sellar or suprasellar involvement was present in 18/57 (31.6%) and 9/57 (15.8%) subjects on computed tomography or magnetic resonance imaging, respectively. The BRAF-positive group was more likely to have adrenal gland involvement (56.2% vs. 20.0%;odds ratio (OR) 5.8, 95% confidence interval (CI) 1.5-17.1, P=0.008), with comparable rates for sellar or suprasellar involvement (28.1% vs. 20.0%;OR 1.6, 95% CI 0.5-5.4, P=0.5) and low random morning cortisol values (37.5% vs. 32.0%;OR 1.3, 95% CI 0.4-3.8, P=0.7). High-sensitivity C-reactive protein was significantly higher in the BRAF V600E-positive group with adrenal involvement when compared to the BRAF V600E-negative group without adrenal involvement (mean concentration: 40 mg/dL vs. 5 mg/dL, P=0.046). HPA axis in ECD tends to involve the adrenal glands in BRAF V600E-positive individuals, without influencing the rates of low cortisol levels, although there is a poor biochemical-radiological concordance in ECD. In this cohort, most subjects with HPA axis involvement did not require glucocorticoid or mineralocorticoid therapy. This study systematically demonstrates an association between the somatic activating BRAF V600E variant and the presence of radiographic and biochemical HPA axis involvement in ECD subjects. TRIAL REGISTRATION: *** Identifier: NCT01417520 © 2020

关键词： Adrenal Insufficiency BRAF Erdheim-Chester Disease Inflammation

来源：评论

学校读者我要写书评

暂无评论

QU-BraTS: MICCAI BraTS 2020 Challenge on Quantifying Uncertainty in Brain Tumor Segmentation - Analysis of Ranking Scores and Benchmarking Results

arXiv

引用

arXiv 2021年

作者： Mehta, Raghav Filos, Angelos Baid, Ujjwal Sako, Chiharu McKinley, Richard Rebsamen, Michael Dätwyler, Katrin Meier, Raphael Radojewski, Piotr Murugesan, Gowtham Krishnan Nalawade, Sahil Ganesh, Chandan Wagner, Ben Yu, Fang F. Fei, Baowei Madhuranthakam, Ananth J. Maldjian, Joseph A. Daza, Laura Gómez, Catalina Arbeláez, Pablo Dai, Chengliang Wang, Shuo Reynaud, Hadrien Mo, Yuanhan Angelini, Elsa Guo, Yike Bai, Wenjia Banerjee, Subhashis Pei, Linmin Murat, A.K. Rosas-González, Sarahi Zemmoura, Ilyess Tauber, Clovis Vu, Minh H. Nyholm, Tufve Löfstedt, Tommy Ballestar, Laura Mora Vilaplana, Veronica McHugh, Hugh Talou, Gonzalo Maso Wang, Alan Patel, Jay Chang, Ken Hoebel, Katharina Gidwani, Mishka Arun, Nishanth Gupta, Sharut Aggarwal, Mehak Singh, Praveer Gerstner, Elizabeth R. Kalpathy-Cramer, Jayashree Boutry, Nicolas Huard, Alexis Vidyaratne, Lasitha Rahman, Md Monibor Iftekharuddin, Khan M. Chazalon, Joseph Puybareau, Elodie Tochon, Guillaume Ma, Jun Cabezas, Mariano Llado, Xavier Oliver, Arnau Valencia, Liliana Valverde, Sergi Amian, Mehdi Soltaninejad, Mohammadreza Myronenko, Andriy Hatamizadeh, Ali Feng, Xue Dou, Quan Tustison, Nicholas Meyer, Craig Shah, Nisarg A. Talbar, Sanjay Weber, Marc-André Mahajan, Abhishek Jakab, Andras Wiest, Roland Fathallah-Shaykh, Hassan M. Nazeri, Arash Milchenko, Mikhail Marcus, Daniel Kotrotsou, Aikaterini Colen, Rivka Freymann, John Kirby, Justin Davatzikos, Christos Menze, Bjoern Bakas, Spyridon Gal, Yarin Arbel, Tal McGill University MontrealQC Canada Group University of Oxford Oxford United Kingdom University of Pennsylvania PhiladelphiaPA United States Department of Radiology Perelman School of Medicine The University of Pennsylvania PhiladelphiaPA United States Department of Pathology and Laboratory Medicine Perelman School of Medicine University of Pennsylvania PhiladelphiaPA United States University Institute of Diagnostic and Interventional Neuroradiology University of Bern Inselspital Bern University Hospital Bern Switzerland Department of Radiology University of Texas Southwestern Medical Center DallasTX United States Department of Bioengineering University of Texas DallasTX United States Advanced Imaging Research Center University of Texas Southwestern Medical Center DallasTX United States Universidad de los Andes Bogotá Colombia Data Science Institute Imperial College London London United Kingdom NIHR Imperial BRC ITMAT Data Science Group Imperial College London London United Kingdom Department of Brain Sciences Imperial College London London United Kingdom Machine Intelligence Unit Indian Statistical Institute Kolkata India Department of CSE University of Calcutta Kolkata India Department of Information Technology Uppsala University Uppsala Sweden Department of Diagnostic Radiology The University of Pittsburgh Medical Center PittsburghPA United States UMR U1253 iBrain Université de Tours Inserm Tours France Department of Radiation Sciences Umeå University Umeå Sweden Department of Computing Science Umeå University Umeå Sweden Signal Theory and Communications Department Universitat Politècnica de Catalunya Barcelona Tech Barcelona Spain Faculty of Medical and Health Sciences University of Auckland Auckland New Zealand Radiology Department Auckland City Hospital Auckland New Zealand Auckland Bioengineering Institute University of Auckland New Zealand Athinoula A. Martinos Center for Biomedical Imaging Department of Radiology

Deep learning (DL) models have provided state-of-the-art performance in various medical imaging benchmarking challenges, including the Brain Tumor Segmentation (BraTS) challenges. However, the task of focal pathology multi-compartment segmentation (e.g., tumor and lesion sub-regions) is particularly challenging, and potential errors hinder translating DL models into clinical workflows. Quantifying the reliability of DL model predictions in the form of uncertainties could enable clinical review of the most uncertain regions, thereby building trust and paving the way toward clinical translation. Several uncertainty estimation methods have recently been introduced for DL medical image segmentation tasks. Developing scores to evaluate and compare the performance of uncertainty measures will assist the end-user in making more informed decisions. In this study, we explore and evaluate a score developed during the BraTS 2019 and BraTS 2020 task on uncertainty quantification (QU-BraTS) and designed to assess and rank uncertainty estimates for brain tumor multi-compartment segmentation. This score (1) rewards uncertainty estimates that produce high confidence in correct assertions and those that assign low confidence levels at incorrect assertions, and (2) penalizes uncertainty measures that lead to a higher percentage of under-confident correct assertions. We further benchmark the segmentation uncertainties generated by 14 independent participating teams of QU-BraTS 2020, all of which also participated in the main BraTS segmentation task. Overall, our findings confirm the importance and complementary value that uncertainty estimates provide to segmentation algorithms, highlighting the need for uncertainty quantification in medical image analyses. Finally, in favor of transparency and reproducibility, our evaluation code is made publicly available at https://***/RagMeh11/QU-BraTS. © 2021, CC BY.

关键词： Benchmarking

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：