City landscapes viewed through home windows influence quality of life, yet perceptions of actual window views at the urban scale remain understudied. This study presents an approach for large-scale mapping of perceptions using 12,334 window view images (WVIs) collected from actual residential properties listed on real estate platforms in Wuhan, China, representing a rarely explored form of urban view imagery that offers advantages over the rendered or simulated window views commonly examined in previous studies. Through a non-immersive virtual reality platform, we collected 27,477 pairwise comparisons across six perceptual dimensions (e.g. preference) from 304 participants based on 499 WVIs. A hybrid neural network model was trained to predict human perceptions of all crowdsourced WVIs and map their spatial distribution. Results reveal significant spatial autocorrelation with distinct hot and cold spots across the whole city. Floor level strongly influences human perceptions: while higher floors offer more preferred and extensive window views, lower-floor windows provide residents with quiet and vivid views. An inference model further shows that window view composition matters considerably: high ratios of sky, trees, and low-rise buildings enhance people’s preferences and perceptions of vividness, whereas high ratios of high-rise buildings increase perceptions of monotony and oppression. Importantly, these effects are non-linear: the excessive presence of certain elements can alter their impact on human perception. This work advances urban-scale understanding of residents’ visual experiences and offers a transferable, human-centric method to inform urban planning and design aimed at improving the visual quality of window views.

Where we live profoundly shapes our health, with urban environments playing a critical role in shaping population health outcomes. As health disparities persist within and between cities, ensuring equitable urban design has become critical to advancing population well-being. Yet most studies focus on case studies of single cities and overlook differences between general, physical and mental health dimensions, limiting our understanding of how urban factors shape health outcomes at scale. Here, to address this gap, we integrate census-tract-level health data, crowdsourced geospatial information and deep learning to identify urban features associated with general, physical and mental health across the most populous urban areas in the United States. Our analysis reveals distinct associative relationships through which urban contextual and socioeconomic factors shape health outcomes. We identify the ranked importance of urban determinants for each health dimension, along with cross-cutting factors that consistently matter. Our findings suggest that urban service enhancements in low-income neighbourhoods are associated with 100–462% greater health gains over high-income areas. Furthermore, we find strong links between the heterogeneity of urban spatial patterns and both health and income inequalities. Overall, our findings highlight strong associations between equitable access to urban services and coherent city planning with observed patterns of population health inequalities across cities.

Characterising and classifying urban functions is a long-standing research focus in urban studies and plays a critical role in urban management and community renewal. However, traditional point-of-interest (POI) categories rely on predefined labels that are often inconsistent across cities and may not fully capture how places are described, represented, or experienced in user-generated data. Further, point-based representations are highly sensitive to spatial aggregation scales, which limits their ability to capture areal functional characteristics and relative differences in POI activity intensity. To address these challenges, we propose a unified framework that, for the first time, leverages place reviews from Google Maps, a form of user-generated geographic information, as a previously untapped POI-linked extended data stream for urban functional inference and classification. Specifically, we employ pre-trained BERT and Vision Transformer models to embed textual and visual information from place reviews, enabling POIs in Singapore and Hong Kong to be represented and clustered within a shared functional embedding space. We then incorporate the weighted volume of place reviews as an indicator of relative POI activity intensity to construct category intensity vectors for spatial units, and demonstrate their effectiveness through cross-city similarity matching tasks. Finally, urban functional classification is conducted across three spatial scales: a 1 km hexagonal grid, administrative areas, and traffic analysis zones (TAZs), using graph neural networks combined with k-means clustering, producing results that preserve spatial continuity and robustness. The proposed framework provides a data-driven approach that highlights the value of place reviews as a complementary data source to conventional POIs and offers a reliable urban functional classification that works across different cities.

Accurately predicting fine-grained urban mobility is essential for optimizing transportation, accessibility, and urban management. However, existing approaches often depend on dynamic data such as trajectories or signaling records, which are sparsely available across cities, thereby limiting their applicability and generalizability to new urban contexts. To address these limitations, this study proposes a Large Model Enhanced Multimodal Representations (LMEMR) framework to learn hourly grid-level mobility dynamics solely from static geospatial data—including remote sensing imagery, building data, street view imagery, and points of interest—which are widely accessible. Large vision–language models are employed to generate natural-language descriptions of each modality, enriching the data with human-understandable semantics. A dual-level contrastive learning strategy aligns raw and textual features both within and across modalities, mitigating semantic gaps and enhancing multimodal consistency. Spatial dependencies are modeled through a graph attention network, and temporal dynamics are captured via a transformer encoder to produce 24-hour mobility sequences. Results from Shenzhen demonstrate that LMEMR outperforms the baseline CLIP model, achieving an $R^{2}$ of 0.856and an 18.07% reduction in MAE. Ablation experiments confirm the effectiveness of semantic enhancement, spatial graph reasoning, and cross-modal fusion. Overall, this research reveals the potential of static multimodal data for dynamic mobility inference, offering a scalable, interpretable, and privacy-friendly solution for smart city planning and management.

Uncrewed Aerial Vehicles (UAVs) have proven to be a transformative technology for the fine-scale perception of the built environment. The recent proliferation of new platforms, sensors, and algorithms creates unprecedented opportunities to understand a complex environment. However, effectively harnessing these opportunities requires a systematic assessment of the emerging methodological practices to address challenges concerning the comparability, reproducibility, and generalizability of the knowledge being produced. Therefore, this study aims to systematically map the dominant methodological workflows in UAV-based built environment perception and critically assess their implications for scientific knowledge production. We conducted a systematic review of 201 peer-reviewed articles in the last decade (2015–2025), complemented by the construction of a novel global dataset of UAV flight policies across 80 countries, to deconstruct the dominant research workflows and to synthesize the progress and challenges across key application domains. Our analysis, leveraging a novel method that integrates PRISMA, machine learning, and Large Language Models, reveals a pronounced convergence in research practices, which stands in contrast to the apparent diversity of available technologies. We determine that the state of the art is characterized by: (i) a geographical concentration of studies in the Global North, correlated with permissive regulatory environments; (ii) a technological path dependency on a ‘standardized toolkit’ of multirotor UAVs and RGB sensors; and (iii) a methodological reliance on self-collected data (91%) that often remains non-public, fostering a research ecosystem of quantitative, computer vision-based analysis. By diagnosing these dominant patterns and their associated challenges, we propose a forward-looking agenda centered on fostering open science, diversifying technologies, and expanding methodological horizons to build a more integrated, robust, and equitable research future.