Search Results (72)

The rapid development of large artificial intelligence (AI) models—large language models (LLMs), multimodel large language models (MLLMs) and vision–language models (VLMs)—has enabled instruction-driven visual understanding, where a single foundation model can recognize and localize arbitrary objects from natural-language prompts. However, predictions from individual models remain inconsistent—LLMs hallucinate nonexistent entities, while VLMs exhibit limited recall and unstable calibration compared to purpose-trained detectors. To address these limitations, a new paradigm termed “multiple large AI model’s consensus” has emerged. In this approach, multiple heterogeneous LLMs, MLLMs or VLMs process a shared visual–textual instruction and generate independent structured outputs (bounding boxes and categories). Next, their results are merged through consensus mechanisms. This cooperative inference improves spatial accuracy and semantic correctness, making it particularly suitable for generating high-quality training datasets for fast real-time object detectors. This survey provides a comprehensive overview of the large multi-AI model’s consensus for object detection. We formalize the concept, review related literature on ensemble reasoning and multimodal perception, and categorize existing methods into four frameworks: prompt-level, reasoning-to-detection, box-level, and hybrid consensus. We further analyze fusion algorithms, evaluation metrics, and benchmark datasets, highlighting their strengths and limitations. Finally, we discuss open challenges—vocabulary alignment, uncertainty calibration, computational efficiency, and bias propagation—and identify emerging trends such as consensus-aware training, structured reasoning, and collaborative perception ecosystems. Full article

We present a psycholinguistic test battery designed to examine the cognitive and affective processes involved in reading Polish poetry. This toolkit combines reader profiling (vocabulary, memory and reading proficiency) with tasks that assess the influence of lexical, textual, affective and poetic features on recognition, context restoration and association generation. Pilot data confirmed the reliability of the measures and their sensitivity to recognised psycholinguistic effects. Vocabulary size and delayed memory rehearsal strongly predicted performance in content restoration, while recognition and association latencies were closely related, indicating shared retrieval mechanisms. Structural and affective properties also influenced responses: line-final words improved recognition but impeded association, with these effects being moderated by word length and frequency. Words that were negatively valenced, abstract and hardly imaginable were restored more accurately than positive or concrete ones. These findings demonstrate the potential of the battery for profiling readers and provide new insights into how Polish poetic language engages memory and associative processes. Full article

Patient stratification strategies based on digital databases and advanced information technology can predict inpatient rehabilitation outcomes and support safe hospital discharge for patients who underwent joint replacement for hip and knee osteoarthritis. The degree of continuity between surgery and rehabilitation, the perioperative process integration, and the setting where rehabilitation is provided are crucial factors to improve care effectiveness, access, minimize readmissions, and cost increase. The primary aim of this scoping review of the literature is to identify perioperative variables that are predictive of inpatient rehabilitation stay after hip and knee arthroplasty for osteoarthritis. These factors are divided by time of assessment through the perioperative pathway and surgical procedure site. The secondary aim is to explore how different data sources and facilities are linked into a patient-centered perioperative pathway. An electronic search of the literature was performed on PubMed, Embase, and Scopus. No time restrictions were applied. All primary research studies investigating predictive factors of inpatient rehabilitation stay after hip and knee osteoarthritis were included. In total, 25 studies were included in the review. Age, caregiver presence, presence of comorbidities, sex, Body Mass Index, Risk Assessment and Prediction Tool composite score, pre-operative Clinician-Reported Outcome Measures, pre-operative Patient-Reported Outcome Measures, and post-operative Barthel Index of autonomy in the Activities of Daily Living were predictive of some degree of inpatient rehabilitation stay in more than one study. The studies were fairly distributed between retrospective and prospective, with multicentric databases more spread among the latter. Data collection occurred in acute hospitals more than in specialized rehabilitation facilities. Using comprehensive models supported by electronic health records and powerful information technologies, analyzing specific inpatient rehabilitation LOS as distinguished from surgical ward rehabilitation, using institutional registries, and including specific rehabilitation factors in these registries, and promoting vocabulary and federated data sharing can strongly enhance the predictivity of models investigating rehabilitation outcomes and support appropriate discharge from inpatient rehabilitation units. Full article

This paper explores the implicit thought of the “body-network” in Maurice Merleau-Ponty’s phenomenology of the body, drawing from both his earlier and later works. It demonstrates that, for Merleau-Ponty, the phenomenal body is inherently interconnected with the world through motor intentionality. Meanwhile, in his later concept of “flesh,” this interconnectedness deepens into a relationship of mutual reflection and chiasmic intertwining, where bodies and the world continuously mirror and permeate each other. The paper then introduces the Huayan Buddhist metaphor of Indra’s Net, along with Fazang’s interpretation of it. A detailed comparative analysis is conducted between Merleau-Pontian body-network and Fazang’s understanding of Indra’s Net. The paper argues for a profound resonance between the primordial characteristics of the Merleau-Pontian body-network—namely, relationality and reflectivity—and Fazang’s key concepts, such as “mutual identity” (相即), “mutual inclusion” (相入), and the contemplative idea that “the images of many bodies are reflected in one mirror” (多身入一鏡像觀). Despite their distinct cultural and philosophical vocabularies, both thinkers construct a relational ontology aimed at deconstructing entrenched dualisms. Through this in-depth comparative study using the Internet of Bodies (IoB) as a case study, this paper demonstrates that the IoB technology exhibits only superficial resemblances to the Merleau-Pontian body-network and Fazang’s interpretation of Indra’s Net. To address the ethical challenges posed by the IoB, it is imperative to integrate the shared philosophical insights of Merleau-Ponty and Fazang in constructing a critical techno-ethics capable of interrogating the ontological reduction and power asymmetries inherent in contemporary technological networks. Merleau-Ponty’s concept of reversible flesh inspires an ethics of contextual sensitivity and user agency, resisting the reduction of lived experience to data points. Meanwhile, Fazang’s Huayan Buddhism, with its principles of mutual identity and mutual inclusion, reveals the relational nature of data, challenging its treatment as neutral or absolute. Together, these philosophies advocate for a decentralized, reciprocal techno-ethics that prioritizes embodied meaning over surveillance and control. Full article

The rapid rise of bio-hybrid robots and hybrid human–AI systems has triggered an explosion of terminology that inhibits clarity and progress. To investigate how terms are defined, we conduct a narrative scoping review and concept analysis. We extract 60 verbatim definitions spanning engineering, human–computer interaction, human factors, biomimetics, philosophy, and policy. Entries are coded on three axes: agency locus (human, shared, machine), integration depth (loose, moderate, high), and normative valence (negative, neutral, positive), and then clustered. Four categories emerged from the analysis: (i) machine-led, low-integration architectures such as neuro-symbolic or “Hybrid-AI” models; (ii) shared, moderately integrated systems like mixed-initiative cobots; (iii) human-led, medium-coupling decision aids; and (iv) human-centric, low-integration frameworks that focus on user agency. Most definitions adopt a generally positive valence, suggesting a gap with risk-heavy popular narratives. We show that, for researchers investigating where living meets machine, terminological precision is more than semantics and it can shape design, accountability, and public trust. This narrative review contributes a comparative taxonomy and a shared lexicon for reporting hybrid systems. Researchers are encouraged to clarify which sense of Hybrid-AI is intended (algorithmic fusion vs. human–AI ensemble), to specify agency locus and integration depth, and to adopt measures consistent with these conceptualizations. Such practices can reduce construct confusion, enhance cross-study comparability, and align design, safety, and regulatory expectations across domains. Full article

This study explores the practice of “soundtalking”, a novel method of participant-led sound practice, across the Dee Estuary in the UK. Over the course of twelve months, the Our Dee Estuary Project facilitated monthly meetings where participants engaged in sound workshops, in-depth discussions, and sound-making activities, with the aim of fostering a deeper connection with both their local and sonic environments. This longitudinal practice-based research study created an environment of sonic learning and listening development, documenting how participants’ interactions and narratives both shape and are shaped by the estuarial environment, its soundscape, and their sense of place. Participant-led conversations formed the basis of the methodology, providing rich qualitative data on how individuals perceive, interpret, and interact with their surroundings and the impact that the soundscape has on the individual. The regular and unstructured discussions revealed the intrinsic value of soundscapes in participants’ lives, emphasising themes of memory, reflection, place attachment, environmental awareness, and well-being. The collaborative nature of the project allowed for the co-creation of a film and a radio soundscape, both of which serve as significant outputs, encapsulating the auditory and emotional essence of the estuary. The study’s initial findings indicate that “soundtalking” as a practice not only enhances participants’ auditory perception but also fosters a sense of community and belonging. The regularity of monthly meetings facilitated the development of a shared acoustic vocabulary and experience among participants, which in turn enriched their collective and individual experiences of the estuary. Soundtalking is proposed as an additional tool in the study of soundscapes to complement and extend more commonly implemented methods, such as soundwalking and soundsitting. Soundtalking demonstrates the efficacy of longitudinal, participant-led approaches in capturing the dynamic and lived experiences of soundscapes and their associated environments, over methods that only create fleeting short-term engagements with the soundscape. In conclusion, the Our Dee Estuary Project demonstrates the transformative potential of soundtalking in deepening our understanding of human–environment interactions and, in addition, has shown that there are both health and well-being aspects that arise from the practice. Beyond this, the project has output a film and a radio sound piece, which not only document but also celebrate the intricate and evolving relationship between the participants and the estuarine soundscape, offering valuable insights for future soundscape research and community engagement initiatives. Full article

This study presents the first systematic evaluation of in-context learning for Tarifit machine translation, a low-resource Amazigh language spoken by 5 million people in Morocco and Europe. We assess three large language models (GPT-4, Claude-3.5, PaLM-2) across Tarifit–Arabic, Tarifit–French, and Tarifit–English translation using 1000 sentence pairs and 5-fold cross-validation. Results show that 8-shot similarity-based demonstration selection achieves optimal performance. GPT-4 achieved 20.2 BLEU for Tarifit–Arabic, 14.8 for Tarifit–French, and 10.9 for Tarifit–English. Linguistic proximity significantly impacts translation quality, with Tarifit–Arabic substantially outperforming other language pairs by 8.4 BLEU points due to shared vocabulary and morphological patterns. Error analysis reveals systematic issues with morphological complexity (42% of errors) and cultural terminology preservation (18% of errors). This work establishes baseline benchmarks for Tarifit translation and demonstrates the viability of in-context learning for morphologically complex low-resource languages, contributing to linguistic equity in AI systems. Full article

Urban flooding presents a growing challenge amid rapid urbanization, climate variability, and fragmented governance. Although simulation and risk assessment tools have advanced considerably, their integration into urban planning remains limited. This study utilized a comprehensive bibliometric analysis of 1293 articles from the Scopus database, selected through a PRISMA-guided workflow, to examine the temporal, structural, and conceptual evolution of simulation, flood risk, and planning in urban flood research from 1980 to 2025. The findings reveal a thematic progression from engineering-centric approaches to broader discourses on resilience, adaptation, and systemic risk. However, disciplinary fragmentation persists, with technical modeling, infrastructure planning, and governance still weakly connected. Despite a shared vocabulary around climate risk and resilience, practical integration into decision-making frameworks remains underdeveloped. The study highlights the need for more cohesive research-practice linkages and calls for frameworks that better align simulation outputs with urban planning imperatives. Full article

The research on teaching scaffolding for preschool children’s oral language development (OLD) has become an important topic in the academic world. However, there remains a lack of evidence-based research on the integration of scaffolding strategies integrated into creative art contexts to support children’s creative expression and language production. In this study, a qualitative case study was conducted to analyze the non-participatory observation and artwork analysis of five-year-old children’s drawing activities in a kindergarten in China based on socio-cultural and scaffolding theories. Three types of core scaffolding strategies were summarized. The findings reveal that the three strategies work together dynamically within the children’s Zone of Proximal Development (ZPD): (1) the visual prompt strategy enriches the vocabulary diversity of metaphors, adjectives, and ordinal words; (2) dialogic narrative co-construction effectively improves narrative coherence across exposition, rising action, climax, and resolution; and (3) emotional engagement strategies foster a safe expressive environment, promoting the integration of affective vocabulary with intrinsic motivation. Accordingly, a three-dimensional integrated “visual-linguistic-emotional” scaffolding model was constructed, emphasizing the practical guidelines of simultaneous scaffolding and gradual scaffolding withdrawal during the warm-up, creation, and sharing sessions of the drawing activity. This study expands the application of scaffolding theory in unstructured art contexts, and provides a systematic practical framework for the design of cross-contextual language support strategies and teacher training in preschool education. Full article

Understanding the mindset of people who die by suicide remains a key research challenge. We map conceptual and emotional word–word co-occurrences in 139 genuine suicide notes and in reference word lists, an Emotional Recall Task, from 200 individuals grouped by high/low depression, anxiety, and stress levels on DASS-21. Positive words cover most of the suicide notes’ vocabulary; however, co-occurrences in suicide notes overlap mostly with those produced by individuals with low anxiety (Jaccard index of 0.42 for valence and 0.38 for arousal). We introduce a “words not said” method: It removes every word that corpus A shares with a comparison corpus B and then checks the emotions of “residual” words in $A-B$. With no leftover emotions, A and B are similar in expressing the same emotions. Simulations indicate this method can classify high/low levels of depression, anxiety and stress with 80% accuracy in a balanced task. After subtracting suicide note words, only the high-anxiety corpus displays no significant residual emotions. Our findings thus pin anxiety as a key latent feature of suicidal psychology and offer an interpretable language-based marker for suicide risk detection. Full article

Current policy recommendations for initial teacher education encourage teaching code-related literacy (phonics, phonological awareness, and phonemic awareness) over pedagogical knowledge, and engaging practice in learning to read. To enhance early childhood pre-service teacher (PST) practices, this mixed-methods pilot study investigated a tool to support PSTs studying birth-to-eight years teaching, pedagogical practice, and knowledge to teach code-related literacy and supplementary vocabulary in conjunction with quality children’s literature. The Non-Scripted Intentional Teaching (N-SIT) tool was developed and then trialled with early childhood PSTs (n = 24) in Queensland, Australia. The participants planned phonics learning experiences using the N-SIT and picture books (e.g., Pig the Pug; Snail and the Whale). Survey data gathered participants’ code-related literacy knowledge before and after the N-SIT training. The data revealed most PSTs felt well-to-somewhat prepared to teach beginning reading and vocabulary and less-to-somewhat prepared to teach phonics. The data further revealed that all participants could define phonics but reported mixed conceptual understandings of phonological and phonemic awareness. The PSTs’ knowledge of phonological awareness, phonemic awareness, and planning for phonics-focused teaching through children’s literature improved post-N-SIT activity. Planned direct systematic phonics instruction strategies through the intentional shared reading of children’s literature and the potential benefits of the N-SIT tool in early childhood initial teacher education are discussed. Full article

Augmented Reality (AR) enables users to see or interact with virtual objects in real-world environments. This case study examines three AR lessons integrated into a beginner-level Spanish course at a university. The participants were 18 undergraduate students enrolled in this class. The AR lessons portrayed a classroom, a coffee shop, and a family setting aimed at improving students’ retention of Spanish vocabulary related to each AR environment. The research data included pre-test and post-test scores, in-class observations, and interviews. Paired-sample t-tests were conducted before and after the AR sessions to examine vocabulary retention. The quantitative findings revealed significant differences in test scores among all three interventions, suggesting the efficacy of AR-based learning methods. A thematic analysis was conducted on qualitative data, encompassing interviews with six students and in-class observations. Students shared that the AR-based lessons made learning more engaging and enhanced vocabulary recall. Student interactions increased, and the AR-based lessons inspired students to transition from technology users to designers. This study addresses the need for more AR learner-centered empirical studies in learning Spanish and provides implications for AR educational application design and implementations. Full article

Research has provided evidence of the protective characteristics of extended family for U.S. emerging adults, but no research has specifically explored cousin relationships. The current study fills this gap by analyzing qualitative data collected from 192 U.S. 18–29-year-old adults (M age = 25.6 years). As this topic is relatively unexplored, examining qualitative data provides scope and vocabulary for further exploration. Participants completed an online survey asking them to describe interactions with extended family, identifying 561 cousins (M age = 28.2 years). A thematic analysis based on the process defined by Braun and Clark generated four primary themes: (1) emerging adults feel varying degrees of closeness and distance with their cousins, (2) relational maintenance with cousins is both planned and incidental, (3) family membership provides resources, and (4) cousins share the same generational position. These results describe important characteristics of the cousin relationship, including moments of unexpected closeness and shared experience of family. The findings also highlight the relevance of sharing a similar life stage within the same family system. Practitioners can utilize findings to help clients identify extended family members that can be tapped for bonding and support. Full article

Research into the unio mystica has revealed what seems to be an area of “real discussion” between scholars of different traditions of mysticism, particularly those of Judaism, Christianity, and Islam. Although this research serves as a promising start to the dialogue among scholars, it has also raised many questions about a “shared moment” that is nevertheless expressed in “irreducibly diverse” and distinct ways in each tradition. What purpose, for instance, can generic cross-cultural categories serve when they mean little or nothing to scholars in each tradition? By contrast, tradition-specific vocabularies are profuse and often difficult to represent in interlinguistic contexts without significant explanation. The challenge of translating mystical texts, imagery, and ideas across cultures and linguistic traditions raises obvious concerns about the misrepresentation and distortion of traditions in an environment of post-colonial critique. Nevertheless, the continued promise of dialogue calls for specialists of these traditions—particularly non-western and non-Christian traditions—to approach, assess, re-formulate, and even challenge the categories of mysticism from within the conceptual and theoretical horizons of the traditions that they research. The present study models such an approach to scholarship in mysticism. It offers a (re)formulation of the unio mystica from within the theoretical frame of the 12th/13th-century Muslim/Sufi mystic, Ibn ʿArabī (d. 638/1240) and early members of his school of thought. By unpacking the primary terms involved in such an account—“God”, the “human being/self”, and “union”—from within the conceptual and theoretical horizons of that tradition, it problematizes the prevailing understanding of the unio mystica constructed from the writings of specialists in Christian mysticism. More importantly, it illustrates the payoff in terms of dialogue (incorporating the critique of existing theories) when each tradition operates confidently from its own milieu, developing its own theoretical resources for mysticism rather than prematurely embracing existing ideas or categories. Full article