The operation and maintenance (O&M) of the communication network supporting a power system are essential for ensuring grid reliability. This paper presents VisualNetO&M, a collaborative visualization system integrated with a digital process twin of the communication network to enhance O&M efficiency. It provides visualizations for key tasks and facilitates collaboration among operators, technicians, and managers. We validated its effectiveness in Xi'an City, China, where it reduced the O&M workflow completion time from 16 hours to just 1 hour. This improvement resulted in a significant economic benefit of nearly 2/3 million USD over 10 months, highlighting the value of VisualNetO&M.
This paper presents an AI-supported system designed to engage visitors with traditional heritage lanterns. Inspired by Song Dynasty-style lanterns, the system guides users through four steps: drawing a lantern shape on a tablet, submitting it to AI generation, carrying the generated digital lantern with a handheld rod, and lighting it up to explore an interactive scroll. When a digital lantern is carried to the scroll, it triggers real-time animations. Observations from our exhibition highlight the system's strong potential for heritage education and cultural engagement in public spaces.
"Prosomoiosi" (Simulation) is a real-time audiovisual study of how successive modeling frameworks overwrite cultural memory. Grounded in media archaeology, the work introduces 'medium alignment', arguing that concepts must be voiced through media that expose their operative logic. A live diffusion pipeline with multi-prompt editing stages the algorithmic politics of selective remembrance, inviting audiences to renegotiate authorship and selfhood at the human-machine frontier through contemplative engagement with the transparent algorithmic process."
Curators face challenges in selecting, ideating, coordinating, and publicizing exhibited works. Recent advances in GenAI has shaped the role of curators in the exhibition design and production process, leading to new adaptations in the exploration, design, planning, and dissemination phases. To explore the impact of GenAI on curatorial practice, we interviewed 13 curators and followed an additional 3 curators in real-world online exhibition production process. We found that GenAI can support theme selection, space design, topic research, artwork selection, and element production. However, GenAI may also stifling critical thinking, perpetuate inaccuracies, leading to fragmented workflows, misperceptions of feasbility, and concerns about copyright. The effectiveness of GenAI tools often depends on the curators understanding of their roles and limitations. This work offer practical design recommendations for curators and suggest future research to design for adopting the use of GenAI in technology-mediated curatorial workflows.
This study examines the effects of two visual guidance techniques, Visual Enhancement and Visual Suppression, on user perception of contextual information in video content. Visual Enhancement introduces explicit visual cues to highlight target content, whereas Visual Suppression attenuates non-target elements, for example, by reducing their brightness. Both approaches aim to isolate specific objects from the background, directing attention to critical information within complex, dynamic scenes. Despite their growing usage, the relative effectiveness of these approaches in guiding attention and their impact on peripheral context awareness remain underexplored. To address this gap, we conducted a controlled user study with 27 participants. The results indicate that Visual Enhancement, through the addition of salient cues, more effectively directs user attention to target information than Visual Suppression. Our findings advance understanding of visual attention in dynamic environments and offer implications for designing visual guidance strategies.
Point clouds are used to measure and assess building processes in areas such as architecture, engineering, and construction. Precise spatial measurements can inform about deviations from a baseline, and the increasing use of sensor data in the context of building information modeling leads to multiple scalar fields containing rich information of individual spatial points. We propose an immersive approach to investigate such multivariate point cloud data from scans of buildings. We compare a switching approach that provides an overview of individual scalar fields with a spotlight that provides local information about all fields simultaneously. Furthermore, locomotion is compared for such immersive analysis with teleportation and with an omnidirectional treadmill. Our results show a preference for free movement and a task-dependence of visualization approaches for the inspection of scalar fields.
This paper introduces a novel framework for human-AI collaborative storytelling using split-screen narrative techniques. Moving beyond conventional AI applications in visual reproduction, we develop a co-creative workflow that integrates computational capabilities with Chinese aesthetic principles to address challenges in temporal and spatial narrative design. Through a case study of The Legend of the White Snake, we demonstrate how AI enhances the symbolic representation of duality, while human expertise ensures cultural authenticity. Key findings include: (1) AI improves efficiency in generating parallel visual sequences and maintaining stylistic consistency; (2) human intervention is essential for contextualizing cultural symbols and emotional resonance; (3) the hybrid approach enables innovative visual storytelling that both respects tradition and expands creative possibilities. This research contributes to culturally-grounded AI applications in digital heritage and narrative innovation.
Human-swarm interaction (HSI) is critical for scalable control of UAV swarm systems. Traditional interfaces struggle with generalization and user workload, especially in immersive environments. Hence, we present ChatHSI, a framework leveraging large language models (LLMs) for swarm task planning. ChatHSI integrates prompt engineering, action validation, and a human-in-the-loop mechanism to improve planning feasibility and executability. We implement ChatHSI in an immersive simulation to improve users' spatial and situational awareness. Our method shows improved task efficiency, reduced workload, and higher usability in user studies. Ablation study proves the effectiveness of prompt context and action validation. The results show the feasibility of LLM-driven interaction for immersive swarm control and point toward adaptive, intuitive, and scalable HSI systems.
The High School Entrance Examination is a critical transitional point in basic education, serving as both a key measure of students' academic levels and a pivotal determinant of their access to senior high school resources and future pathways. Analyzing the link between mock exam scores and senior high school preference choices is essential for scientifically guiding school entrance planning and refining preparation strategies. This study uses a genetic algorithm to optimize random forest parameters and presents AdmPredVis, a visualization system with multi-view collaboration for admission prediction. The system includes three core views: Prediction results view visualizes educational progression, score distributions, and ZY rankings via junior-senior high school tables; Model explanation view visualizes random forest decision processes and individual attribution; Student list supports detailed instance information display, data editing, and real-time prediction feedback to provide guidance for students. Case analyses and expert evaluations validate the model's effectiveness and AdmPredVis's practical value in educational planning.
Rising concerns about interpretable AI systems have boosted demand for explainable AI (XAI) and visualization interfaces. While model-agnostic local methods that explain the rationale behind a certain prediction are widely used, existing visualization systems either employ a single explanation method or examine only one prediction at a time, limiting analysis of model behavior from multiple perspectives. We propose a matrix-based XAI visualization method that enables users to choose and overlay two explanation types from a set of heterogeneous local explanations. Feature-based, rule-based, and counterfactual explanations can be toggled or super-imposed in the same layout, enabling rapid, multi-facet inspection of model behavior. To measure users' comprehension and interpretation of a large-scale visualization of multiple explanations, we introduce XAI-VLAT, a task set adapted from the Visualization Literacy Assessment Test (VLAT). We include consistency and proxy (simulation) tasks that probe reasoning with paired explanations. We conducted a user study by applying XAI-VLAT to our proposed visualization. Accuracy was high across most tasks; the rule-plus-counterfactual pair achieved the best results, whereas color-dominant visualization for feature-based explanations sometimes drew attention away from other cues. Qualitative feedback shows that a unified overlay helps them cross-check explanations and simulate model behavior.
This paper proposes an AI-assisted speculative narrative design workflow as a critical tool to address and respond to ecological ethics and anthropocentrism in the Anthropocene. Speculating on a post-climate collapse, ocean-dominated future, Project Serum constructs a multi-species narrative through the lens of a deep-sea court trial, where humans, whales, and robots contest ecological justice and species dominance. Combining speculative design methodologies with multispecies world building, this work builds a narrative blueprint using a structured 5W1H framework and AI-enhanced story design. This blueprint is then remediated into multiple formats, including video, postcards, and comic posters, through the proposed AI-assisted workflows integrating AI image generation and human editorial control. By applying these narrative remediation strategies with AI, this study challenges anthropocentric narrative structures and offers a reusable workflow for multi-modal, multi-species storytelling. It demonstrates the speculative narrative's potential as a reflective intervention, capable of destabilising normative narrative hierarchies and foregrounding ecological justice with AI-mediated production.
In the District Heating (DH) sector, the analysis and monitoring of data from DH substations is crucial to keeping the entire DH network running efficiently. Clustering of DH substations based on multivariate data helps analyze their behavior over time. In this context, a visualization-based analysis approach can be particularly beneficial. In this paper, we explore the use of dynamic hypergraph visualization to analyze the clustering results of DH network substations over time. We present the initial results of designing and implementing a visual analytics dashboard that supports DH experts in analyzing different behaviors of DH substations. In the proposed dashboard, we adopt the Parallel Aggregated Ordered Hypergraph (PAOH) technique to visualize dynamic hypergraphs, which provides a compact visualization of multivariate data clustering over time. Moreover, we include additional views with complementary visualizations supporting the analysis and understanding of the dynamic hypergraph main view. We showcase the capability of our dashboard applied on a real DH dataset.
We translate the time-based viewing logic of Chinese handscrolls - scatter perspective, paced unfolding, and qiyun (spirit resonance) - into two complementary outputs: a silk-printed long scroll and a three-screen walking installation. A period-aware training pipeline fine-tunes diffusion models with dynasty-specific adapters and a structural controller that guides blank space and large-scale composition. Rather than isolated images, we assemble a continuous digital scroll and stage it at architectural scale so that viewers literally "travel" through the landscape. A small expert review and an audience study suggest high stylistic fidelity and a strong sense of embodied flow. We argue that the digital form extends-rather than replaces-the paper original, offering a practical model for reactivating classical image logics in contemporary immersive media.
The contemporary dance piece Dark Side of the Moon is inspired by a fantasy novel and explores the relationship between AI and humans. Its fantasy-based theme presents the challenge of balancing creative expression through art with the need to maintain a coherent narrative in the performance. To address this challenge, a co-creation process was conducted between a dance team and a technical team, with a focus on the design and integration of visual feedback. Through observations of the co-creation process, interviews and discussions with the dancers, as well as comments collected from the audience through questionnaires, we found that visual feedback supported both dancer creativity and audience interpretation of the narrative. It is surprising to find that some visual information that limited the dancers' movements helped the audience better understand the performance's theme. This revealed the dual nature of visual feedback as both a potential constraint and an enhancement, which offers valuable insights for future explorations on how to balance creative expression with making performances more accessible to the audience.
In this paper, we describe a visual analytics tool for monitoring and controlling water quality focusing on the Rhine River. We investigated correlations between several water quality metrics in a static and dynamic way. The results of the analysis are visually represented on a dashboard that allows interactive navigation in the data to support the confirmation, rejection, or refinement of the data-related hypotheses. We integrated easy-to-understand visualization techniques for non-experts in visualization in a multiple coordinated view. We illustrate the usefulness of the approach by applying it to water quality data at a measurement station called Weil am Rhein. We focus on data quality metrics including temperature, pH values, oxygen content, and the electric conductivity while we found a major correlation between temperature and oxygen saturation. Finally, we discuss scalability issues and limitations also as some kind of performance analysis before we conclude the paper with future directions.
This study investigated whether emotional responses to graph color schemes differ between sexes. We conducted an online evaluation using bar charts with fourteen color schemes. The analysis revealed that some schemes induced large sex differences, whereas others induced small differences. Using entropy, we grouped the schemes by emotional concentration, revealing four categories that combined high and low concentrations with large and small sex differences. These two aspects appeared to be independent, although unified features for each group were not identified. These findings highlight the potential of emotion-aware visualization designs that consider individual differences.
Although virtual production (VP) offers cinematic and immersive storytelling by aligning a high-end camera with multiple LED screens through a central server, the creation of high-quality 3D scenery with real-time interactions remains complex and resource-intensive. This paper explores the potential of expanding cinematic virtual production scenes through three innovative AI-driven approaches: (1) AI-generated 360 degrees panoramas from text prompts to produce immersive backgrounds; (2) Direct text/AI-generated image-to-3D environment conversion using mesh generation; (3) An end-to-end AI pipeline for rapid stylized scene construction. We demonstrate each workflow through real-time avatar-interactive shooting scenarios. Our approach bridges technical and artistic domains and aims to show how these AI-driven workflows could accelerate scene creation, enable novel cinematic experiences, and reveal the future potential of visual content generation.
Effective visualization is essential for cultural heritage interpretation. However, existing visualization systems remain constrained by fragmented data integration and limited exploration capabilities for multimodal heritage data. This paper presents HeritageExplorer, an interactive system that synergizes large language models (LLMs) with dynamic visualizations to enable progressive heritage exploration. Our approach constructs a comprehensive knowledge graph integrating 831 historic buildings in Guangzhou, which unifies their architectural, spatial, temporal, and contextual attributes. The system's novel integration of KG-enhanced contextual understanding with LLMs supports: natural language query understanding and seamless coupling with interactive visualizations. Quantitative evaluation demonstrates consistent improvements in factual accuracy across heritage tasks, while case studies illustrate its successful application in diverse exploration scenarios.
Electric power is the foundation of modern society, yet Europe is currently facing an energy crisis, increasing interest in power generation, energy infrastructure, and grid resilience. However, power plant data are complex and multidimensional, making it difficult to gain an overview or understanding. Visualization methods can help to reduce cognitive load and facilitate exploration of such data. In this paper, we propose EuroEnergyVis, a web-based visualization approach designed for the interactive exploration of power plant data across European countries. The design requirements were motivated by gaps identified in prior work. We conducted interviews with six domain experts in power systems and energy, which indicate that our tool enhances the user experience when exploring European power plants. Their reflections also suggest directions for future work.
"Quantum est in libris" explores the intersection of the archaic and the modern. On one side, there are manuscript materials from the Estonian National Museum's more than century-old archive, describing the life experiences of Estonian people; on the other side is technology that transforms these materials into a dynamic and interactive experience. Connecting technology and cultural heritage is the visitor, who turns texts into inputs for a screen sculpture. Historical narratives come to life visually through contemporary technological language. Because the employed text-to-video AI models by Runway Gen-3 and Gen-4 have not previously interacted with Estonian heritage, we see how machines today "read the world." "Quantum est in libris" introduces an exciting yet unsettling new dimension to the concept of cultural heritage: in a world where data is fluid and interpretations unstable, heritage status becomes fragile. In the digital environment, issues related to heritage are no longer just about preservation and transmission, but also about media representation, machine creativity, and interpretive error. Who or what shapes memory processes and memory spaces-and how?
