Abstract: Interactive explanatory visualization of volume data in immersive environments has always been challenging, as depth information causes severe visual confusion and occlusion, while XR headset controllers or other interactive devices are difficult to achieve fine-grained interaction. In this report, we will introduce a tool for automatically constructing immersive science communication animations for scientific volume data, which only requires users to interactively interpret one or two 2D slices of volume data (with the help of mature image processing algorithms). The interpretive visualization results on 2D slices will be automatically migrated to 3D immersive space, avoiding users' direct interactive interpretation in 3D space. Compared with state-of-the-art image-to-image style transfer neural networks, the proposed method can be constructed on desktop PCs without time-consuming training. We will also briefly introduce how to automatically construct science communication animations for multi-scale scientific data.

Speaker Bio: Richen Liu, Associate Professor, received his PhD from Peking University in 2017. His research interests include intelligent virtual-real fusion and visualization, intelligent multimedia. He has led 2 NSFC projects and multiple departmental projects, and participated in multiple National Social Science Foundation key projects and general projects. In the past four years, he has published 15 papers as first author or corresponding author in CCF A/B class, ACM/IEEE Trans, and Chinese Academy of Sciences Zone 1/2 journals, including ACM CHI, ACM MM, IEEE TVCG/THMS/TBD, etc. He has won first prize in teaching achievement, first prize in undergraduate excellent teaching award. In the past four years, he has guided students to win 3 Jiangsu Province excellent graduation thesis awards (including 1 provincial first prize), 4 Jiangsu Province graduate practice innovation plan projects, and more than 20 national awards in mainstream competitions recognized by the Ministry of Education.

Abstract: With the continuous development of virtual reality, augmented reality, and mixed reality technologies, visualization environments with different degrees of virtuality provide unprecedented possibilities for scientific data display and interaction. Compared with traditional 2D interfaces, this type of cross-reality visualization can more intuitively present complex 3D structures, multi-scale hierarchies, and dynamic evolution processes, significantly improving users' understanding and exploration efficiency of scientific data. This report will introduce the visualization and interaction methods we designed in environments with different degrees of virtuality, covering key strategies such as immersive presentation, multimodal interaction, spatial mapping, and task adaptation, aiming to provide more adaptive and scalable interaction solutions for future-oriented scientific data analysis.

Speaker Bio: Lingyun Yu, Associate Professor, doctoral supervisor, head of the Interactive Visualization Research Group, mainly engaged in immersive visualization, extended reality, and human-computer interaction. Has published over 80 papers in visualization and human-computer interaction fields, and has led National Natural Science Foundation projects (1 general project, 1 youth project). As first author, won the IEEE VIS Scientific Visualization Twelve-Year Test of Time Award and Best Paper Nomination Award, and has served as paper chair multiple times at top conferences in the field.

Abstract: This report will explore how to design and implement mixed reality technology with contextual visualization functions for medical education and training under interdisciplinary research methodology. Specifically, I will introduce three research areas: inner ear anatomy teaching, traditional Chinese medicine prescription teaching, and personal protective equipment training. In these works, we fully utilized the characteristics of mixed reality such as virtual-real integration, spatial immersion, and natural interaction, proposing innovative interactive teaching and training methods. The effectiveness of these methods has been confirmed through a series of randomized controlled trials and expert feedback.

Speaker Bio: Liang Zhou, Associate Researcher, Assistant Professor, doctoral supervisor, working at the National Institute of Health Data Science, Peking University. Research interests include visualization, health data visual analytics, and mixed reality.

Abstract: In complex system analysis fields such as weather forecasting, oceanographic research, and disaster warning, ensemble data is widely used. However, how to effectively express the diverse uncertainties contained in ensemble data remains a severe challenge in current visualization research. Currently, visualization of ensemble data mainly relies on two-dimensional display methods, which are limited by visual channels to some extent, easily causing cognitive bias, data confusion, and information hiding. This report aims to explore ensemble data visualization methods based on stereoscopic imaging technology, effectively utilizing depth cues to enhance visual encoding and expression capabilities for multivariate information, and improve perception and reasoning of uncertainty distribution.

Speaker Bio: Shizhou Zhang, doctoral supervisor at the School of Computer Science, Northwestern Polytechnical University, tenured Associate Professor, selected for Shaanxi Province Science and Technology Rising Star and Northwestern Polytechnical University "Soaring Star" talent support program. Received bachelor's and doctoral degrees from Xi'an Jiaotong University in 2010 and 2017 respectively. Research interests include computer vision and image processing. In recent years, he has been dedicated to developing autonomous evolutionary learning methods and applying them to UAV visual tasks and air-ground collaborative visual perception tasks to improve UAV visual perception capabilities. Related research has published over 60 papers in international top journals IEEE TIP, TMM, TCSVT, TGRS, T-ITS, Pattern Recognition and top conferences CVPR, ICCV, ECCV, NeurIPS, ICML, AAAI, IJCAI, ACM MM, etc. One paper was selected as ACM MM HCMA 2023 Best Paper Award, and one paper was selected as ESI highly cited paper. He has led key engineering projects, National Natural Science Foundation projects, Shaanxi Province Natural Science Development Fund, and participated as a technical backbone in National Natural Science Foundation key projects, National Key R&D Program, etc. Project R&D results have been successfully applied to multiple series of UAV systems.