Search Results (308)

While creative Artificial Intelligence (AI) tools offer unprecedented creative power, their outputs often create a “personalization gap” by converging towards a generalized “average aesthetic” that ignores nuanced user preferences. This study addresses this challenge with a proof-of-concept computational framework to model and predict subjective color choices, aiming to make creative systems more human-centered. Our dual-track methodology attempts to decouple user preference into “inherent aesthetic profiles” and “contextual design decisions.” Through a dual-level study with 111 participants, we quantified inherent aesthetics into a vector library and trained a Gradient Boosting Decision Tree (GBDT) model on contextual data to predict design choices. The model achieved a predictive accuracy of 40.8%, and a grouped permutation importance analysis revealed the Product Category (Importance = 0.416) as the dominant predictor, providing evidence that design context is paramount. Crucially, a subsequent exploratory user validation study, analyzed with a linear mixed-effects model, showed our personalized recommendations were rated as significantly more satisfying (β = 1.278, p < 0.001) than those of a non-personalized baseline. This research provides a foundational framework for modeling subjective preference by distinguishing between stable traits and dynamic choices, offering a potential pathway to steer creative AI beyond generic outputs towards more personal and context-aware creative partners. Full article

There has been a surge in interest in and implementation of motion capture (MoCap)-based lessons in animation, creative education, and performance training, leading to an increasing number of studies on this topic. While recent studies have summarized these developments, few have been conducted that synthesize existing findings into a theoretical framework. Building upon the Cognitive Affective Model of Immersive Learning (CAMIL), this study proposes the Cognitive Affective Model of Motion Capture Training (CAMMT) as a theoretical and research-based framework for explaining how MoCap fosters creative cognition in computer animation practice. The model identifies six affective and cognitive constructs: Control and Active Learning, Reflective Thinking, Perceptual Motor Skills, Emotional Expressive, Artistic Innovation, and Collaborative Construction that describe how MoCap’s technological affordances of immersion and interactivity support creativity in animation practice. The findings indicate that instructional and design methods from less immersive media can be effectively adapted to MoCap environments. Although originally developed for animation education, CAMMT contributes to broader theories of creative design processes by linking cognitive, affective, and performative dimensions of embodied interaction. This study offers guidance for researchers and designers exploring creative and embodied interaction across digital performance and design contexts. Full article

Over the past five years, generative AI has carved out a major role in architecture, especially in education and visual idea generation. Most of the time, the literature talks about AI as a tool, an assistant, or sometimes a co-creator, always highlighting efficiency and the end product in architectural design. There is a steady rise in empirical studies, yet the real impact on how young architects learn still lacks a solid theory behind it. In this systematic review, we dig into peer-reviewed work from 2015 to 2025, looking at how generative AI fits into architectural design education. Using PRISMA guidelines, we pull together findings from 40 papers across architecture, design studies, human–computer interaction and educational research. What stands out is a clear tension: on one hand, students crank out more creative work; on the other, their reflective engagement drops, especially when AI steps in as a replacement during early ideation instead of working alongside them. To address this, we introduce the idea of “AI sparring”. Here, generative AI is not just a helper—it becomes a provocateur, pushing students to think critically and develop stronger architectural concepts. Our review offers new ways to interpret AI’s role, moving beyond seeing it just as a productivity booster. Instead, we argue for AI as an active, reflective partner in education, and we lay out practical recommendations for studio-based teaching and future research. This paper is a theoretical review and conceptual proposal, and we urge future studies to test these ideas in practice. Full article

Incorporating STEM (Science, Technology, Engineering, and Mathematics) into early childhood education has been associated with children’s holistic development. STEM education not only enhances critical thinking, creativity, problem-solving, and other 21st-century skills but also contributes significantly to cognitive growth, emotional regulation, and social abilities. Within the early childhood context, the use of play and toys emerges as a natural and powerful medium for introducing STEM concepts in developmentally appropriate and engaging ways. Play and toys have a prominent role, and previous studies have provided strong evidence on their educational benefits. Toys enhanced with technological characteristics (Technology-Enhanced Toys—TETs), such as coding and interactive toys, are increasingly being viewed as cultural tools that mediate learning and nurture cognitive and collaborative skills among young learners. However, the impact TETs have on young children’s STEM learning remains largely unexplored. This qualitative observational study, grounded in a socio-cultural perspective, explored how 37 children aged 3 to 4 years in four early childhood settings in Greece exhibited STEM-related behaviours during free play with technology-enhanced toys. Data were collected through systematic video recordings and written observations over a three-month period that involved interacting with various TETs, such as Bee-Bot, Coko Robot, a remote-controlled dog, and others. Results indicate that playing with TETs enhanced problem-solving, computational thinking, and collaboration, thus affirming the positive influence of digital technology and the potential of TETs to enrich early STEM education. Implications for equity, the importance of teachers’ professional development in effectively integrating TETs into early childhood curricula and the need for further research will also be discussed. Full article

The application of Generative Artificial Intelligence (GenAI)—defined as a class of AI systems capable of autonomously generating new content such as images, texts, and design solutions based on learned data patterns—has become increasingly widespread in creative design. By supporting ideation, rapid trial-and-error, and data-driven decision-making, GenAI enables designers to explore design alternatives more efficiently and enhances human–computer interaction experiences. In design practice, GenAI functions not only as a productivity-enhancing tool but also as a collaborative partner that assists users in visual exploration, concept refinement, and iterative development. However, users still face a certain learning curve before effectively adopting these technologies. Within the framework of human-centered artificial intelligence, contemporary design practices place greater emphasis on inclusivity across diverse user groups and on enabling intuitive “what-you-think-is-what-you-get” interaction experiences. From a sustainable design perspective, GenAI’s capabilities in digital simulation, rapid iteration, and automated feedback contribute to more efficient design workflows, reduced collaboration costs, and broader access to creative participation for users with varying levels of expertise. These characteristics play a crucial role in enhancing the accessibility of design resources and supporting the long-term sustainability of creative processes. Focusing on the context of China’s digital design industry, this study investigates the application of GenAI in design workflows through an empirical case study of Zhitu AI, a generative design tool developed by Beijing Didi Infinity Technology Development Co., Ltd. The study conducts a literature review to outline the role of GenAI in visual design processes and employs observation-based experiments and semi-structured interviews with users of varying levels of design expertise. The findings reveal key pain points across stages such as prompt formulation, secondary editing, and asset generation. Drawing on the Kano model, the study further identifies potential design opportunities and discusses their value in improving efficiency, supporting non-expert users, and promoting more sustainable and inclusive design practices. Full article

Artificial intelligence is reshaping approaches to architectural heritage conservation by enabling a deeper understanding of how people perceive and experience historic built environments. This study employs deep learning and large language models (LLMs) to explore public perceptions of the Qinghefang Historical and Cultural District in Hangzhou, illustrating how AI-driven analytics can inform intelligent heritage management and architectural revitalization. Large-scale public online reviews were processed through BERTopic-based clustering to extract thematic structures of experience, while interpretive synthesis was refined using an LLM to identify core perceptual dimensions including Hangzhou Housing & Residential Choice, Hangzhou Urban Tourism & Culture, Hangzhou Food & Dining, and Qinghefang Culture & Creative. Sentiment polarity and emotional intensity were quantified using a fine-tuned BERT model, revealing distinct affective and perceptual patterns across the district’s architectural and cultural spaces. The results demonstrate that AI-based textual analytics can effectively decode human–heritage interactions, offering actionable insights for data-informed conservation, visitors’ experience optimization, and sustainable management of historic districts. This research contributes to the emerging field of AI-driven innovation in architectural heritage by bridging computational intelligence and heritage conservation practice. Full article

Research on extracting building from airborne point clouds is abundant, yet discussions regarding scenarios where vegetation and building structures are closely intertwined with similar height in rural areas remain relatively scarce. This thesis adopts a region representative of typical rural building features in China as an experimental site to conduct research on building classification procedures from airborne point clouds. Firstly, the multi-level grid size is dynamically determined through slope analysis to creatively segment and recognize terrain type, then differentiated filtering parameters are applied to various terrains to fully extract ground points, providing a ground reference for building classification. Secondly, the selection of building Region of Interest is conducted by multiple geometric feature differences between building and other objects based on watershed segmentation results, which eliminates interference from non-building points, significantly reducing redundant and unnecessary mathematical computation. Finally, refined building classification is achieved based on multiple morphological differences between buildings and other objects. The experimental results show that the precision, recall, and F1 of both datasets exceeded 93.37%, 97.05%, and 95.17%, respectively. The average precision, recall, and F1 reached 94.02%, 97.20%, and 95.58%, respectively. This method demonstrates successful building classification in rural areas, showing strong adaptability and practicality for the extraction of various building data. Full article

The Extended Digital Narrative (XDN) research project explores how experimental creative practice with emerging technologies generates critical insights into algorithmic narrativity—the intersection of human narrative understanding and computational data processing. This article presents five case studies demonstrating that direct engagement with AI and Extended Reality platforms is essential for humanities research on new genres of digital storytelling. Lina Harder’s Hedy Lamar Chatbot examines how generative AI chatbots construct historical personas, revealing biases in training data and platform constraints. Scott Rettberg’s Republicans in Love investigates text-to-image generation as a writing environment for political satire, documenting rapid changes in AI aesthetics and content moderation. David Jhave Johnston’s Messages to Humanity demonstrates how Runway’s Act-One enables solo filmmaking, collapsing traditional production hierarchies. Haoyuan Tang’s video game project reframes LLM integration by prioritizing player actions over dialogue, challenging assumptions about AI’s role in interactive narratives. Sérgio Galvão Roxo’s Her Name Was Gisberta employs Virtual Reality for social education against transphobia, utilizing perspective-taking techniques for empathy development. These projects demonstrate that practice-based research is not merely artistic production but a vital methodology for understanding how AI and XR platforms shape—and are shaped by—human narrative capacities. Full article

Automata and artificial intelligence (AI) have long occupied a central place in cultural and artistic imagination, and the recent proliferation of AI-generated artworks has intensified debates about authorship, creativity, and human agency. Empirical studies show that audiences often perceive AI-generated works as less authentic or emotionally resonant than human creations, with authorship attribution strongly shaping esthetic judgments. Yet little attention has been paid to how AI systems themselves evaluate creative authorship. This study investigates how large language models (LLMs) evaluate literary quality under different framings of authorship—Human, AI, or Human+AI collaboration. Using a questionnaire-based experimental design, we prompted four instruction-tuned LLMs (ChatGPT 4, Gemini 2, Gemma 3, and LLaMA 3) to read and assess three short stories in Italian, originally generated by ChatGPT 4 in the narrative style of Roald Dahl. For each story × authorship condition × model combination, we collected 100 questionnaire completions, yielding 3600 responses in total. Across esthetic, literary, and inclusiveness dimensions, the stated authorship systematically conditioned model judgments: identical stories were consistently rated more favorably when framed as human-authored or human–AI co-authored than when labeled as AI-authored, revealing a robust negative bias toward AI authorship. Model-specific analyses further indicate distinctive evaluative profiles and inclusiveness thresholds across proprietary and open-source systems. Our findings extend research on attribution bias into the computational realm, showing that LLM-based evaluations reproduce human-like assumptions about creative agency and literary value. We publicly release all materials to facilitate transparency and future comparative work on AI-mediated literary evaluation. Full article

Text-to-image (T2I) generation, a core component of generative artificial intelligence(AI), is increasingly important for creative industries and human–computer interaction. Despite impressive progress in realism and diversity, diffusion models still exhibit critical security blind spots particularly in the Transformer key-value mapping mechanism that underpins cross-modal alignment. Existing backdoor attacks often rely on large-scale data poisoning or extensive fine-tuning, leading to low efficiency and limited stealth. To address these challenges, we propose two efficient backdoor attack methods AttnBackdoor and SemBackdoor grounded in the Transformer’s key-value storage principle. AttnBackdoor injects precise mappings between trigger prompts and target instances by fine-tuning the key-value projection matrices in U-Net cross-attention layers (≈5% of parameters). SemBackdoor establishes semantic-level mappings by editing the text encoder’s MLP projection matrix (≈0.3% of parameters). Both approaches achieve high attack success rates (>90%), with SemBackdoor reaching 98.6% and AttnBackdoor 97.2%. They also reduce parameter updates and training time by 1–2 orders of magnitude compared to prior work while preserving benign generation quality. Our findings reveal dual vulnerabilities at visual and semantic levels and provide a foundation for developing next generation defenses for secure generative AI. Full article

Objectives: Lifestyle behaviors such as physical activity, cognitive engagement, social interaction, diet, sleep, and vascular risk management are increasingly recognized as contributors to cognitive aging and dementia risk. Although many middle-aged and older adults express interest in maintaining brain health, less is known about their beliefs about brain-healthy behaviors or their preferences for receiving multicomponent brain health interventions. This study examined adults’ ratings of the usefulness of a wide range of lifestyle activities for brain health and their preferred formats for receiving support. Methods: A 60-item online survey was administered to compensated volunteers aged 40 years and older through a commercial provider. The questionnaire assessed perceived usefulness of lifestyle-based brain health activities and preferred intervention delivery formats. The analytic sample included 761 respondents. Descriptive statistics were computed for all ratings and differences by age group and gender were tested using MANOVA with post hoc comparisons adjusted for multiple testing. Results: Participants endorsed many lifestyle activities as helpful for brain health. Mentally stimulating activities, good sleep, stress management, and creative activities received the highest ratings, whereas strength training, meditation, language learning, and computer-based cognitive training were rated lower. Aerobic exercise and mentally stimulating activities were most frequently selected as the single most important activity. Significant effects of age, gender, and their interaction were observed, with younger men and older women generally rating activities more favorably. With respect to desire for services, over half of participants preferred receiving a cognitive assessment, and many favored online education or app-based tools. Conclusions: Middle-aged and older adults recognize a wide range of lifestyle factors as potentially beneficial for brain health and express strong interest in structured support, particularly assessments and digital resources. These findings can inform the design of flexible, multicomponent brain health interventions aligned with adults’ preferences and priorities. Full article

The construction industry remains one of the main contributors to environmental degradation due to its high material consumption and massive waste generation. This study introduces Granizzo, a hybrid methodological framework that integrates artificial intelligence (AI), parametric design, and digital fabrication to transform construction and demolition waste (CDW) into sustainable architectural mosaics. The workflow involves material selection, AI-driven classification of fragments, generative design algorithms for pattern optimization, and CNC-based experimental prototyping. A dataset comprising brick, cement, marble, glass, and stone fragments was analyzed using a Random Forest classifier, achieving an average accuracy above 90%. Parametric design algorithms based on circle packing and tessellation achieved up to 92% surface coverage, reducing voids and optimizing formal diversity compared to manually assembled mosaics. Prototypes fabricated with CNC molds exhibited 35% shorter assembly times and 20% fewer voids, confirming the technical feasibility of the proposed process. A preliminary Life Cycle Assessment (LCA) revealed measurable environmental benefits in energy savings and CO₂ reduction. The findings suggest that Granizzo constitutes a replicable methodological platform that merges digital precision and sustainable materiality, enabling a circular approach to architectural production and aligning with contemporary challenges of design innovation, material reuse, and computational creativity. Full article

This study investigates the perceptions of Greek sixth-grade students regarding Artificial Intelligence (AI). Understanding students’ pre-instructional conceptions is essential for developing targeted interventions that build on existing knowledge rather than assuming conceptual deficits. A qualitative design was employed with 229 students from seven elementary schools in Athens, Greece. Data were collected through open-ended questions and word association tasks, then analyzed using Walan’s AI perceptions framework as an integrated set of analytical lenses (cognitive, affective, behavioral/use, and ethical considerations). Findings revealed that students hold multifaceted conceptions of AI. Cognitively, they described AI as robots, computational systems, software tools, and autonomous learning programs. Affectively, they expressed ambivalence, balancing appreciation of AI’s usefulness with concerns over potential risks. Behaviorally, they identified interactive question–answer functions, creative applications, and everyday assistance roles. Ethically, students raised issues of responsible use, societal implications, and human–AI relationships. This study contributes to international research, highlighting that primary students’ understandings of AI are more nuanced than is sometimes assumed, and offer empirical insights for designing culturally responsive, ethically informed AI literacy curricula. Full article

While large language models (LLMs) such as ChatGPT, Claude, and DeepSeek are evaluated based on their accuracy and truthfulness, “hallucinations” betray underlying structural limitations. These results are not simply incorrect answers, but statistical resonances; they are instances where models stabilize into statistically significant (though semantically unfounded) response patterns. Current frameworks fail to accommodate contextual semantics, experiential time, and intentionality as key dimensions for effective experience-based decision-making in complex digital spaces. This article presents an integration paradigm offered by the theory of uncertainty and incompleteness of information, extended by the Sophimatics approach with 2D complex time (t = t + i·t₀) and Super Time Cognitive Neural Network (STCNN) that provides both memory management, imagination enhancement, and creativity generation as computational primitives. By integrating probability with plausibility, credibility, and possibility, our model reconsiders the issue of evaluating the reliability of LLM results as a problem that goes beyond traditional probabilistic approaches. Accepting that hallucinations are an emerging phenomenon of resonance between statistical distributions, we suggest an extended probability method in which these resonances can be mitigated and directed towards a coherent cognitive understanding. The paper places this approach in the broader perspective of digital transformation at the information systems level and its implications for AI reliability, explainability, and adaptive decision-making in post-generative AI. Intuitive scenarios are described, based on the inclusion of complex time and Sophimatics in theoretical modelling, illustrating how prediction, historical-contextual adoption, and resistance to paradoxical or contradictory information are strengthened. The results point to this paradigm as a springboard for reliable, human-aligned AI capable of enabling digital transformation in sectors such as healthcare, finance, and governance. Full article

Motion capture (MoCap) is increasingly recognized as a powerful multimodal immersive learning technology, providing embodied interaction and real-time motion visualization that enrich educational experiences. Although MoCap is gaining prominence within educational research, its pedagogical value and integration into computer animation training environments have received relatively limited systematic investigation. This review synthesizes findings from 17 studies to analyze how MoCap supports instructional design, creative development, and workflow efficiency in animation education. Results show that MoCap enables a multimodal learning process by combining visual, kinesthetic, and performative modalities, strengthening learners’ sense of presence, agency, and perceptual–motor understanding. Furthermore, we identified five key technical affordances of MoCap, including precision and fidelity, multi-actor and creative control, interactivity and immersion, perceptual–motor learning, and emotional expressiveness, which together shape both cognitive and creative learning outcomes. Emerging trends highlight MoCap’s growing convergence with VR/AR, XR, real-time rendering engines, and AI-augmented motion analysis, expanding its role in the design of immersive and interactive educational systems. This review offers insights into the use of MoCap in animation education research and provides a springboard for future work on more immersive and industry-relevant training. Full article