This article examines how hacker culture, often conceptualized as immaterial and virtual, is in fact materially and spatially constituted through its entanglement with physical places. Focusing on Las Vegas and DEF CON, this article shows the emergence of hackerspectacle, a place-bound mode of interfacing that enables the dual-direction seepage of form and power: subcultural acts leave material residues in policies and design, while the city's spectacle economy filters back to script hackers' style, memory, and self-understanding. The article traces how a three-decade coupling between DEF CON and Las Vegas co-produces both the conference and the city. By intervening in hotel systems, accessing controls, and displaying infrastructures, hackers appropriate Las Vegas's visual language and spatial affordances to craft their placed identity. Conceptually, this case advances STS discussions on the materiality of digital cultures. Empirically, it shows a city-level co-construction. The article also diagnoses a drift from subversion to absorption as DEF CON mirrors Las Vegas's streamlining, commercialization, and surveillance. The article is based on original archival research, ethnographic work, and media analysis. It draws on DEF CON programs, hacker zines, public and anonymized interviews, news coverage, and visual materials, and it situates hacker practices within Las Vegas's legal, architectural, and economic history. It also offers a generalizable template for studying how technocultures take place, literally, and will interest readers of infrastructure studies, digital materialities, urban technopolitics, and the socio-spatial dynamics of subcultures.

Background/Objectives: To evaluate the diagnostic accuracy of two visual large language models (vLLMs), GPT-4o (OpenAI) and Claude Sonnet 3.5 (Anthropic), for detecting brain metastases in routine MRI using combined imaging and textual input. Methods: This retrospective study included 31 patients with and 46 without brain metastases with underlying melanoma (n = 24), lung cancer (n = 23), breast cancer (n = 17), or renal cell carcinoma (n = 13). In total, 100 MRI examinations (50 with, 50 without metastases) were provided to both vLLMs using a single representative slice per sequence, together with clinical history and the referring question. The generated free-text reports were evaluated for detection accuracy, overdiagnosis, correct sequence recognition, anatomical localization, lesion laterality, and lesion size estimation. Results: Both vLLMs showed perfect sensitivity (100% for both) but very low specificity (GPT-4o: 8%, Sonnet 3.5: 4%; p = 0.625), resulting in low diagnostic accuracy (GPT-4o: 54%, Sonnet 3.5: 52%; p = 0.625). Sequence identification was highly accurate in both models, with GPT-4o performing significantly better (100% vs. 93%; p < 0.05). Identification of the anatomical brain region (70% vs. 72%; p = 1.00) and lesion laterality (62% vs. 76%; p = 0.189) was comparable. Both models hallucinated additional lesions in 12% of cases. Lesion size measurements showed no significant differences between the models or in comparison with the radiologist. Conclusions: GPT-4o and Claude Sonnet 3.5 can generate radiological reports and detect brain metastases with excellent sensitivity, but their very low specificity, frequent hallucinations, and limited spatial reliability currently preclude clinical application. Future work should address how the balance between visual and textual input influences diagnostic behavior in vLLMs.

VQA explanation task aims to explain the decision-making process of VQA models in a way that is easily understandable to humans. Existing methods mostly use visual location or natural language explanation approaches to generate corresponding rationales. Although significant progress has been made, these frameworks are bottlenecked by the following challenges: 1) Uni-modal paradigm inevitably leads to semantic ambiguity of explanations. 2) The reasoning process cannot be faithfully responded to and suffers from logical inconsistency. 3) Human-annotated explanations are expensive and time-consuming to collect. In this paper, we introduce a new Semi-supervised VQA Multi-modal Explanation (SME) method via self-critical learning, which addresses the above challenges by leveraging both visual and textual explanations to comprehensively reveal the inference process of the model. Meanwhile, in order to improve the logical consistency between answers and rationales, we design a novel self-critical strategy to evaluate candidate explanations based on answer reward scores. More importantly, our method can benefit from a tremendous amount of samples without human-annotated explanations with semi-supervised learning. Extensive automatic measures and human evaluations all show the effectiveness of our method. Finally, the framework achieves a new state-of-the-art performance on the three VQA explanation datasets. The code for this work is publicly available at https://github.com/Fake10086/MM-Explanations.

Limited cross-modal activation to visual language in the auditory cortices of hearing-impaired children

Image captioning task requires effective combination of visual feature extraction and natural language generation. This study compares four pre-trained models such as Vision Transformer (ViT-B/16), ResNet-18, VGG-16, and DenseNet-121 when applied as frozen feature extractors in a prefix-based captioning framework using a partially trainable BERT-baseuncased text encoder. Experiments were conducted on a 32,000-image subset of the MS COCO 2017 captions dataset (28,000 training, 4,000 validation) under a limited training environment. Performance was evaluated using training cross-entropy loss. Results show DenseNet-121 achieved the lowest final loss (0.2894), followed by VGG-16 (0.3198), ResNet-18 (0.3935), and ViTB/16 (0.7002). DenseNet-121 demonstrated superior feature richness and fastest generalization, while ViT exhibited slowest convergence. These findings suggest that, under resource-constrained scenarios with frozen backbones, DenseNet-121 is the most effective choice among the evaluated architectures.

Visual communication in modern Indian paintings relies heavily on symbols to convey meanings that go beyond simple or literal representation. These symbols come from various sources, such as cultural memories, mythology, personal gender experiences, social and political conditions, and the personal stories of artists. Together, these elements create a strong visual language through which modern Indian art shares ideas, emotions, and social issues. This paper examines semiotics as a tool for interpreting symbols in visual communication, specifically in modern Indian paintings. The study builds on the semiotic theories of Ferdinand de Saussure and Charles Sanders Peirce, focusing on how visual signs create meaning through representation and interpretation. Using a qualitative research approach, the study conducts semiotic analysis on selected modern Indian artworks to explore the symbolic roles of colour, form, line, space, and imagery. Artists like S. H. Raza, V. S. Gaitonde and F. N. Souza use abstraction and expressive distortion to express themes of spirituality, identity, and existential struggle. Meanwhile, women artists such as Arpita Singh, Gogi Saroj Pal, and Nalini Malani use symbolic imagery to tackle issues of gender, memory, mythology, and social critique. Their artworks show how personal experiences and shared cultural symbols intersect, creating layered and complex visual meanings within modern Indian visual communication. The study highlights the viewer's active role in meaning-making, where interpretation is shaped by cultural background, visual awareness, and personal experiences. The findings indicate that semiotics provides a clear and effective framework for understanding the intricate symbolic systems found in modern Indian paintings. By placing modern Indian art within a semiotic and visual communication context, this research contributes to interdisciplinary discussions and underscores the continuing importance of symbolic interpretation in understanding contemporary visual expression at both national and global levels.

Generative artificial intelligence (AI) is transforming audiovisual production, yet empirical research on Gulf media ecosystems is limited. This study examines how generative AI reconfigures visual language in the Kuwaiti context, conceptualized as shifts in aesthetic conventions, stylistic patterning, and symbolic repertoires of audiovisual materials, rather than in narrative structures or production workflows, alongside audience negotiations of credibility regarding synthetic media. We constructed a multi-layered corpus of publicly accessible videos, audience comments, and metadata drawn from Kuwaiti audiovisual platforms and applied a Python-based computational research design. Visual change was operationalized using the Shot Dynamics Index (SDI), which captures pacing and editing rhythms, and the AI-Visual Index (AVI), measuring the prevalence of AI-associated visual cues. These visual measures were integrated with audience discourse analysis, including a Credibility Lexicon Score (CLS), topic modeling, sentiment analysis, and network-based diffusion, community, and centrality metrics. Descriptive fixed-effects models link these analytical layers without making causal claims and are supported by extensive robustness checks. The results showed that a sustained increase in AVI, partially decoupled from pacing (SDI), was accompanied by intensified verification discourse (higher CLS) and clustering around AI-tagged content within central network hubs and cross-platform bridging nodes. The study contributes cross-cultural evidence on algorithmic aesthetics and advances transparent, transferable measurement frameworks, highlighting provenance labeling and dialect-aware NLP as viable mechanisms for supporting credibility in AI-mediated audiovisual environments.

The present research utilizes three popular multimodal AI systems (Gemini, Cloud AI, ChatGPT) to evaluate their ability to interpret visual language by analyzing responses to three different images. In addition to evaluating the systems' ability to produce accurate and detailed descriptions of each image, this research evaluated their ability to contextualize each description within an appropriate framework of understanding, as well as assess their response times. Results indicate that ChatGPT provided the most accurate and descriptive descriptions of the images analyzed in this study, particularly those which depicted emotionally and/or socially nuanced scenes; that Gemini performed reasonably well in terms of conceptual interpretation, though was inconsistent in its provision of specific details regarding the image(s); and that while Cloud AI responded more quickly than either ChatGPT or Gemini, it failed to provide as much detail or relevance to the situation presented in the images. These findings emphasize the need to develop multimodal AI systems that balance speed, emotional intelligence, and semantic accuracy to be used in the real world when reasoning with images.

This paper examines an underexplored area in fashion scholarship by focusing on how prosthetic limbs are esthetically integrated as forms of identity expression and cultural authorship. Using visual and discourse analysis of content from the Chinese social media platform RedNote (Xiaohongshu), the study explores how disabled influencers style and incorporate prosthetic limbs into fashion practices and online self-presentation within digital visual culture. Drawing from the affirmative model of disability and fashion theory, the paper uses Fashioning Affirmative Embodiment as a conceptual lens that synthesizes existing frameworks to analyze prosthetic customization as aesthetic agency, subcultural participation, and resistance to ableist norms. While the findings are not intended to generalize across disabled populations, this study offers a situated analysis of visual practices that challenge and broaden the global scope of fashion and disability studies. The findings point to the development of a visual language in which prosthetic limbs are styled in relation to trend-driven and genre-based esthetics. In these practices, assistive technologies become associated with narrative meaning and stylistic experimentation. By focusing on the Chinese digital context, the study contributes to existing discussions in disability and fashion studies by situating embodiment, belonging, and self-expression beyond predominantly Western frameworks.

Indigenous muralists across the Pacific have adopted urban art aesthetics as a strategic means of asserting ongoing presence, celebrating cultural traditions, and articulating visions of Indigenous futures. This research note examines two murals by Hawaiian artists Carl F.K. Pao, Cory Taum, and Solomon Enos that were included in the 2021 Bishop Museum exhibition POW! WOW! The First Decade: From Hawai‘i to the World. Through urban art’s accessible visual language, these artists assert an enduring Indigenous will to self-define, ground their work in ancestral knowledge, and articulate temporal visions that span past, present, and multiple futures.

Accurate non-invasive diagnosis of early-stage ovarian cancer remains challenging because of the limited number of biomarkers. Although artificial intelligence algorithms show promise for ovarian cancer diagnosis, their reliance on specialized engineering knowledge hinders their accessibility. The recent emergence of visual large language models such as GPT-4o has further expanded the potential of AI in this domain. GPT-4o was trained to automatically recognize ovarian lesions, report key computed tomography (CT) features of ovarian lesions, and make a benign or malignant diagnosis based on these features. Radiologists and gynecologic oncologists independently reviewed the GPT-4o reports and evaluated GPT-4o's performance. GPT-4o achieved diagnostic accuracies of 80.80%, 79.14%, and 93.33% in the three datasets. Its performance surpassed that of gynecologic oncologist with 10years of experience but was inferior to that of gynecologic oncologist with 16years of experience and radiologists with ≥7 years of experience. The clinician-rated reliability in detecting the four key CT features was 4.22/5.00 for cyst wall and septum status; 4.24/5.00 for nodular or papillary protrusions; 4.30/5.00 for density and enhancement distribution; and 4.25/5.00 for cystic-solid characteristics. The use of GPT-4o increased the accuracy of radiologist and gynecologic oncologist diagnoses by 1.96% and 10.50%, respectively. GPT-4o identifies the key CT features of ovarian cancer and achieves promising diagnostic accuracy with high-quality diagnostic evidence.

Abstract. The integration of digital technologies aimed at the knowledge of the existing built heritage is now a well-established practice in the scientific field. The objective remains the optimization of methodologies intended to coordinate different needs, ranging from the integration of technical and documentary aspects to the enhancement of communication and dissemination of knowledge of heritage assets. While Heritage Building Information Modelling (HBIM) represents the most widespread methodology for Facility Management (FM), enabling control over the entire life cycle of buildings, the challenge lies in integrating Extended Reality (XR) tools, which allow operators to consult, update, and interact immersively with digital models, expanding their potential both for technical management and for heritage valorization. In this direction, the contribution proposes a workflow that integrates Augmented Reality (AR) into Heritage Building Information Modelling (HBIM) processes, enabling real-time updating of information through on-site observations. The case study concerns the complex of Santa Maria del Rifugio in Naples (Italy), characterized by historical stratifications from the 15th to the 20th century and by complex structural and conservation conditions. Starting from the digital survey, the information model was enriched with dimensional and analytical metadata, synchronized in AR through the structuring of a Visual Programming Language (VPL) algorithm that promotes the overlap between real and virtual through Image Targets, interaction with data stored in the BIM Common Data Environment (CDE), and their real-time updating. The result is an interactive system that enhances understanding, management, and valorization of cultural heritage, offering an engaging cognitive and immersive representation of architecture.

Abstract. The paper presents a two-phase methodological protocol for creating an Extended Matrix (EM) based BIM model (EM-BIM). The EM, developed by the ISPC-CNR institute, was designed to manage data, metadata, and paradata related to virtual reconstructions in archaeology. This research introduces a methodology for geometric and informational modelling, based on an HBIM approach, of existing architectural heritage using EM tools within Historical Building Information Modelling (HBIM) approach. Knowledge-based structure of EM is similar to object-based system of BIM: both methodologies are grounded in a relational approach to 3D data structuring. Furthermore, EM explicitly incorporates paradata using a visual knowledge graph system. Tested on the Doric Stoa of Priene, the first phase of the study focused on decoding the graphic language of EM to automatically generate and populate ad hoc descriptors within the HBIM environment using a visual programming language (VPL). A second phase involved the setting up of a BuildingSMART Data Dictionary (bsDD) to map the EM classification against the IFC classes, to standardise the generation of dedicated ‘property sets’.

The role of public transportation in Ibero-America is significant in terms of economic efficiency and urban social cohesion. It possesses the capacity to attenuate disparities in health, employment, educational attainment, and recreational engagement. Furthermore, public transportation can serve as a catalyst for the formation of a collective identity among its users. This study methodically examines the relationship between the visual rhetorical language employed by public transportation systems in Ibero-America and the values associated with these media. The study is founded on the hypothesis that public transportation embodies fundamental principles such as equal opportunities, equity, accessibility, safety, and sustainability, which are essential for serving the most disadvantaged sectors. The research seeks to identify the extent to which these values are iconographically manifested in the brands of public transport systems, with the purpose of evaluating the coherence between the values that support these services and their visual representation. This study uses a methodology that combines qualitative and quantitative analysis to examine sixty-eight public transport companies in Ibero-America. The aim was to determine how the values of public transport are expressed in iconic, graphic, and visual ways and whether transport companies try to reflect these values in their branding. This study contributes to academic knowledge on visual rhetoric in public transport and provides valuable insights for effective brand design and management. The findings can be used to promote the consideration of universal transport values in brands, thereby contributing to more effective planning and management of public transport in the region.

Visual language models show widespread visual deficits on neuropsychological tests

Keringkam is a traditional Sarawak Malay embroidery with gold or silver wrapped threads. The embroidered motifs function as visual symbols deeply embedded in both Islamic art and local cultural identity. This study employs Ferdinand de Saussure’s semiotic theory to analyze how the motifs in keringkam operate as signs, wherein the visual forms (signifiers) correspond to culturally constructed meanings (signifieds) derived from Islamic aesthetic principles and Sarawak Malay social traditions. By examining recurring floral, vegetal, geometric, and cosmological patterns, the research demonstrates how keringkam functions as a visual language that conveys local values. Drawing on primary data collected from Sarawak embroiderers and cultural practitioners, together with secondary literature on Islamic ornamentation and Malay material culture, this study challenges conventional interpretations that regard keringkam as merely decorative. Instead, it positions keringkam motifs as symbolic expressions of faith, identity, and heritage, thereby contributing to broader scholarly discussions on Islamic textile semiotics and Malay visual culture.

This paper examines how sociology can think cinematically through the creation of sociological film, achieved by blending sociological and filmic imaginations. While traditional visual sociology often treats film as a tool for data collection, this study argues that cinema functions as a medium of thought that constructs and communicates social knowledge through image, rhythm and affect. Drawing on Just Black? (1992), Talking Heads (1980), Anything Can Happen (1995) and Chronicle of a Summer (1961), the paper demonstrates how cinematic form can embody key dimensions of sociological imagination, such as lived experience, structure–actor relations and reflexivity. By analysing how interviews and framing translate sociological inquiry into visual language, it highlights film's potential to generate affective, collaborative and public modes of understanding. The study concludes that sociological film transforms the camera from an instrument of observation into a participant in thought, making the sociological imagination visible and experiential.

Abstract: This paper examines how Nilima Sheikh employs visual metaphors to construct layered narratives that engage deeply with memory, tradition, gendered experience, displacement, and socio-political critique. Drawing upon traditional South Asian art forms, literary references, and personal histories, Sheikh’s metaphoric strategies generate a poetic visual language that negotiates between the personal and the political. Through a critical analysis of selected bodies of work — particularly Each Night Put Kashmir in Your Dreams — this study situates her practice at the intersection of narrative painting, cultural memory, and political commentary. Supported by artist interviews, archival material, and exhibition literature, the paper argues that Sheikh’s metaphoric structures operate not simply as aesthetic devices, but as epistemological tools that reveal hidden histories, fractured geographies, and gendered modes of witnessing within contemporary Indian art. Keywords: Nilima Sheikh, visual metaphor, memory, contemporary Indian art, narrative painting, Kashmir, displacement, feminist aesthetics, tradition and modernity, socio-political art, intertextuality

This critical literature review examines how African American English (AAE) and American Sign Language (ASL)-including Black ASL-shape literacy development in Black Deaf students. It situates these linguistic systems within broader historical and contemporary contexts of literacy acquisition, highlighting systemic inequities in educational practices that marginalize nonstandard dialects and visual languages. Understanding these dynamics is essential to developing more inclusive and effective literacy instruction that honors the linguistic identities of Black Deaf learners. Drawing from sociocultural theory, (socio)cognitive theory, poststructuralism, and critical race theory, this literature review employs a thematic synthesis to analyze how these frameworks illuminate the complex interplay between language, identity, and literacy development among Black Deaf students. The review integrates empirical and theoretical studies across hearing and Deaf populations, situating findings within broader historical and societal literacy trends. It also addresses contemporary challenges such as multilingualism and the role of nonpublic educational settings in shaping literacy outcomes. The findings reveal that prevailing educational systems frequently rely on deficit-based models that marginalize AAE and ASL, thereby limiting literacy opportunities for Black Deaf students. The literature advocates for a shift toward diversity-affirming frameworks that recognize AAE and ASL as vital cognitive and cultural assets supporting literacy acquisition. Key themes include the importance of culturally responsive pedagogy and inclusive educational policies that validate students' linguistic identities, dismantle systemic biases, and foster equitable literacy development. These insights underscore the need for educational practices that integrate students' home languages and social contexts to promote academic success. Affirming AAE and ASL in educational settings is essential for promoting equitable literacy development. The review recommends integrating these theoretical perspectives into policy reform and within the fields of education and speech-language pathology. Future research should explore the roles of family, community, and multimodal literacy practices in supporting Black Deaf learners across diverse educational contexts.

Visual Large Language Models Exhibit Human-Level Cognitive Flexibility in the Wisconsin Card Sorting Test