“CSIG图像图形中国行”是由中国图象图形学学会主办的学术活动,旨在推动图像图形学科的普及,加强各高校研究所以及企业间的交流。自2017年4月起至今,分别在大连理工、哈尔滨工业大学、兰州大学、杭州电子科技大学、北京交通大学、海南大学、福州大学、西北农林科技大学等成功举办了100余期,线上线下参会人数累计数万人次,受到业界好评。本次“CSIG图像图形中国行”由可视化与可视分析专业委员会与复旦大学大数据学院联合承办,交流主题为“大数据可视化与智能人机交互”。
活动信息
主办单位:中国图象图形学学会(CSIG)
承办单位:CSIG可视化与可视分析专委会、复旦大学大数据学院
时间:2025年12月6日13:30-17:30
地点:复旦大学邯郸校区光华楼东主楼2601室
执行主席
复旦大学 陈思明 研究员
简介:陈思明,复旦大学大数据学院青年研究员,博士生导师,上海市高层次引进人才,复旦大学可视分析与智能决策实验室负责人(FDUVIS),复旦仲英学者。曾任德国弗劳恩霍夫智能分析和信息系统研究所(Fraunhofer IAIS)研究科学家与德国波恩大学的博士后研究员。复旦学士(2011)、北大博士(2017)。从事大数据可视化与可视分析,交互式人工智能的研究,主要研究方向包括:AI+VIS、大模型驱动的可视分析、社交媒体分析、自动驾驶、金融科技、数字孪生等,共发表论文100余篇,其中在IEEE VIS,IEEE TVCG, ACM CHI, CSCW,UIST等顶级国际人机交互人智协同会议以及期刊(CCF A)上发表40余篇文章。曾获评AI2000十年间国际可视化研究最有影响力提名奖(全球100名),主持、参与国家、省部级项目十余项,担任IEEE VIS 国际程序委员会委员,IEEE CG&A国际期刊副主编,Visual Informatics期刊青年编委、IEEE PacificVis论文(VizNotes)主席,ChinaVis论文主席及数据分析挑战赛主席,VGI Geovisual Analytics Workshop地理时空可视分析研讨会共同主席等。更多信息请访问http://fduvis.net。
华东师范大学 李晨辉 副教授
简介:李晨辉,华东师范大学计算机科学与技术学院副教授,博士生导师,CSIG可视化与可视分析专委会委员,CCF计算机辅助设计与图形学专委会执行委员。博士毕业于香港理工大学,研究方向包括数据可视化、计算机图形学、智能艺术设计等,主持国家自然科学基金面上项目、国家重点研发计划项目子课题、上海市自然科学基金项目、华为委托项目等十余项,在IEEE VIS、IEEE TVCG、NeurIPS、CVPR、ICML、ACM CHI等国际会议及期刊上发表学术论文50余篇,其中CCF A或Trans.论文30余篇;担任VCIBA国际期刊青年编委、IEEE VIS等国际会议PC;6次担任国内学术会议论文主席、组织主席,曾获2次国际学术会议最佳论文奖,曾获2020年度上海市科技进步特等奖、2022年度上海市高等教育教学成果二等奖、2023年全国高校优秀美育案例、2024年度上海计算机学会科学技术奖二等奖。
特邀专家
孙国道 浙江工业大学 教授
讲者简介:孙国道,博士,浙江工业大学计算机科学与技术学院教授,博士生导师。国家优秀青年科学基金获得者,浙江省杰出青年基金获得者,入选浙江省高校领军人才培养计划(青年优秀人才)。主要从事大数据挖掘、可视化和可视分析方面的研究,聚焦时空定位、自然语言和视频影像等代表性数据。在可视化领域相关期刊和会议发表论文40余篇(第一作者/通讯作者30余篇),包括IEEE TVCG、IEEE VIS、IEEE TMM、ACM TIST等,并担任IEEE VIS、IEEE Pacific VIS、VINCI、ChinaVis等国际学术会议的程序委员会委员和多个国内外著名期刊和会议的审稿人。主持国家自然科学基金面上项目、青年基金、国家重点研发计划子课题、浙江省重点研发课题、浙江省自然科学基金一般项目等项目,作为项目技术骨干参与国家重点研发计划和国家自然科学基金重点项目。荣获浙江省科技进步一等奖(多源异构时空定位数据的关联分析平台及应用)。
报告题目: Intelligent Visual Analysis of Multimodal Data
报告摘要: One of the current hot and challenging topics in big data research lies in the intelligent representation, correlation analysis, and understanding of data. Interactive visual analytics combines the domain reasoning capabilities of expert users with traditional data mining to help synthesize and distill information, validate expected patterns, and discover unknown phenomena. This talk will explore research on “visual analytics + artificial intelligence” for representative data types such as video imagery, multidimensional tabular data, and spatio-temporal positioning data, and will introduce our team’s visual analytics work and representative case studies in these areas, with specific application scenarios including video surveillance, business intelligence, and urban transportation.
马昱欣 南方科技大学 副教授
讲者简介:马昱欣,南方科技大学计算机科学与工程系副教授、研究员,于浙江大学计算机科学与技术学院获得学士与博士学位,曾任美国亚利桑那州立大学VADER实验室博士后。主要研究方向为数据可视化、交互式数据分析、人机交互,包括基于数据可视化方法的人工智能可解释性研究、高维数据和时空数据的可视分析方法、结合智能计算的交互式教学和创意设计系统等。目前发表论文四十余篇,在IEEE TVCG、IEEE VIS、ACM CHI等国际顶级期刊会议上发表十余篇长文,曾获得ACM CHI最佳论文提名、CVMJ期刊年度最佳论文提名、陆增镛CAD&CG高科技奖等奖项,以及参与国内首部数据可视化专著的编写工作。
报告题目: Visual Analytics on Explainable AI: Case Studies on Analyzing Optimization Processes and Language Models
报告摘要: In recent years, the widespread application of AI techniques has enabled the inspection, understanding, and prediction of the behaviors of individuals and groups within massive datasets, providing new insights into society and the world while optimizing social governance. However, the black-box nature of complex algorithms and models has limited users' understanding of the mechanisms and predictions of the models, thereby affecting trust in the models. This talk will address the challenges by exploring how visual analytics approaches can facilitate the comprehension of complex algorithms and models, with a particular emphasis on optimization processes and language models.
唐谈 浙江大学 研究员
讲者简介:唐谈,浙江大学艺术与考古学院"百人计划"研究员(博士生导师),在人机交互/可视化/虚拟现实等方向国际顶级学术会议和期刊(如IEEE TVCG, IEEE VIS/VR, ACM CHI/UIST)上发表论文30余篇,并多次获最佳论文提名奖,作为"中国历代绘画大系"展览策划组成员参与第60届威尼斯国际艺术双年展中国国家馆项目,相关工作获得了央视等权威媒体多次报道。
报告题目: Intelligent Visual Analytics of Social Media Big Data
报告摘要: In recent years, cross-platform information diffusion has gradually become a research hotspot. However, due to the unclear nature of implicit propagation paths, existing studies often perform poorly on downstream tasks such as rumor source tracing and diffusion prediction. This talk focuses on how to leverage visual analytics and human-computer interaction techniques to overcome computational bottlenecks in cross-platform entity association. Based on a causal inference framework, we reconstruct cross-platform diffusion networks to support event correlation and source tracing, thereby enabling the development of a trustworthy and visualizable system for monitoring information dissemination.
李权 上海科技大学 助理教授
讲者简介:李权,上海科技大学信息科学与技术学院助理教授(终身教授序列)、研究员、博士生导师,从事人工智能及可视分析、可解释性机器学习以及人机交互技术的研究。他博士毕业于香港科技大学计算机科学与工程学系。任中国图象图形学学会可视化与可视分析专委会委员,IEEE VIS Paper程序委员会委员、ChinaVis论文国际程序委员会委员、IEEE VIS, EuroVis, PacificVis, ChinaVis, ACM CHI/CSCW及TVCG等顶级学术会议期刊审稿人,他曾任美国佐治亚理工学院计算机科学与工程学院的访问研究员、微众银行人工智能部资深研究员及网易游戏资深研究员。他的学术成果发表在IEEE VIS, EuroVis, IEEE PacificVis, ACM CHI, CSCW, UIST, IUI, CGF, TVCG等可视化及人机交互顶级期刊和会议,获得ACM CHI 2025最佳论文奖,并主持国家自然科学基金面上项目。更多信息见https://faculty.sist.shanghaitech.edu.cn/liquan/
报告题目: More Than a Tool: Weaving AI into the Fabric of Compassionate Care
报告摘要: What if AI could do more than just automate tasks? What if it could become a thoughtful partner in the delicate ecosystem of care? Our work explores this future, designing AI systems that integrate deeply into the human side of medicine. We begin by tackling the silent challenges: the immense effort behind creating quality data, and the hidden biases within AI models. Our systems transform these foundational steps from solitary chores into collaborative dialogues, ensuring the technology is built on a bedrock of accuracy and fairness. But a reliable tool is just the beginning. We then show how this thoughtful AI can step into the most human of spaces. It can act as an invisible scaffold, supporting communication for neurodiverse individuals. More profoundly, it can serve as a dynamic teaching companion, unlocking the wisdom of past medical records to cultivate the diagnostic intuition of the next generation of doctors. This is not about replacing the clinician. It's about thoughtfully extending their capabilities—from the raw data up, through the clinic, and into the classroom—to weave a future of healthcare that is not only smarter and fairer, but ultimately more human.
朱倩 中国人民大学 讲师
讲者简介:朱倩,博士,2024年于香港科技大学计算机科学与工程系获博士学位,2025年加入中国人民大学数据工程与知识工程教育部重点实验室并任讲师,现为中国人民大学吴玉章青年英才,CCF人机交互专委会执行委员。
她的研究方向为人机交互、数据可视化和虚拟/增强现实(VR/AR),发表十余篇CCF-A类期刊/会议论文。主持国家自然科学基金青年科学基金项目(C类),担任ACM CHI,IEEE Pacific Vis程序委员会委员和多个人机交互、数据可视化领域期刊/会议的审稿人。
报告题目: Visualization and Interaction Research for Spatial Contexts
报告摘要: With the development of virtual and augmented reality technologies, human-computer interaction and data visualization systems are gradually expanding from traditional desktop and mobile platforms into embodied, context-aware three-dimensional spaces. In this context, this talk will systematically present, through multiple immersive data interaction and analysis prototype systems and empirical studies, how humans can perceive, understand, and act on data when it is no longer confined to a two-dimensional screen but embedded in a tangible and manipulable 3D space. Specifically, the presentation will share several recent studies in immersive data visualization, focusing on: how to leverage digital twins to construct spatialized data storytelling systems, enabling users to move from “understanding data” to “experiencing data”; how to design effective multimodal visual interaction techniques in real-world scenarios to naturally support everyday decision-making and complex analysis; and how to build multi-view composite visualizations in 3D space to support more flexible, autonomous, and embodied immersive analysis scenarios. Looking ahead, with the deep integration of large models and XR technologies, the talk envisions a next-generation visualization system in which humans can experience intelligence, proactivity, and interactivity in three-dimensional space—truly transforming the way people comprehend and engage with complex information.
张宇 华为 研究员
讲者简介:张宇,华为基础软件创新实验室研究员。博士毕业于牛津大学计算机系,本科毕业于北京大学智能科学系。在人机交互与可视化领域开展工作,主要研究方向为交互式机器学习、智能用户界面以及数字人文,相关论文发表于IEEE VIS、ACM CHI、IEEE TVCG、ACM TIIS等会议与期刊。曾获IEEE PacificVis Journal Track最佳论文提名。
报告题目: Intelligence Amplification in Human-Data Interaction
报告摘要: The rise of data-driven sciences and the recent surge of generative AI have contributed to an overwhelming flood of data, making it increasingly critical to amplify both experts' and casual users' capabilities to interact with data. In this talk, I will present our work on this challenge, spanning two stages of human-data interaction: data preparation and data analysis. For data preparation, in the context of digital humanities, we developed AI-assisted data labeling systems and generalized the underlying labeling workflow into a visual programming toolkit for customizing labeling workflows. For data analysis, we designed adaptive visual analysis systems powered by generative AI to support different analytical tasks and task stages. Finally, I will introduce interaction modeling as an approach for evaluating and optimizing human-data interaction processes.
张玮 浙大城市学院 讲师
讲者简介:张玮,浙大城市学院智能文化计算研究中心教师。博士毕业于浙江大学,专注可视分析、数字人文研究。在IEEE VIS、IEEE TVCG、NeurIPS等国际会议及期刊上发表学术论文20余篇。
报告题目: Exploring Historical Narratives: When Visual Analytics Meets Digital Humanities
报告摘要: Visual analytics has become a core technology driving the development of digital humanities, enabling efficient data analysis and the discovery of new insights within vast historical databases, thereby greatly enriching the understanding of such data. This talk will introduce how interactive visual analytics techniques can be used to explore biographical data of historical figures and reveal social structures and hidden historical narratives from a group perspective. The presentation will further extend the analytical lens from historical figures to cultural and artistic collections, capturing the multidimensional historical contexts represented by these artifacts. Through the analysis of well-documented traditional Chinese handscroll paintings, the talk will demonstrate how static artworks can be transformed into vivid historical narratives.
会议时间
13:00 - 13:20 会议签到
13:20 - 13:30 领导致辞
13:30 - 14:00 主题报告:孙国道教授
14:00 - 14:30 主题报告:马昱欣副教授
14:30 - 15:00 主题报告:唐谈研究员
15:00 - 15:30 茶歇
15:30 - 16:00 主题报告:李权研究员
16:00 - 16:30 主题报告:朱倩助理教授
16:30 - 17:00 主题报告:张宇研究员
17:00 - 17:30 主题报告:张玮助理教授
联系方式
承办方联系人:
陈老师 simingchen3@gmail.com
主办方联系人:
王老师 010-82544754 info@csig.org.cn
Copyright © 2025 中国图象图形学学会 京公网安备 11010802035643号 京ICP备12009057号-1
地址:北京市海淀区中关村东路95号 邮编:100190