2026-10-25

7 arXiv categories· 96 journals· 2530 candidates — 7 worth your time· 113 without an open abstract

Headline

Road lane quality assessment from egocentric dashcam video

Today, together

tag shift
no tag ran above its 30-day average
shares references
Does telehealth narrow the urban-rural gap? A dual assessment of in-person and telehealth prenatal care accessibility shares 6 references with Mapping regulatory frictions in telehealth: A spatial accessibility approach using the R-2SVCA model (2026-10-09)
3 of them
  • Measures of Spatial Accessibility to Health Care in a GIS Environment: Synthesis and a Case Study in the Chicago Region
  • An enhanced two-step floating catchment area (E2SFCA) method for measuring spatial accessibility to primary care physicians
  • Incorporating broadband durability in measuring geographic access to health care in the era of telehealth: A case example of the 2-step virtual catchment area (2SVCA) Method
; Does telehealth narrow the urban-rural gap? A dual assessment of in-person and telehealth prenatal care accessibility shares 3 references with Spatial Inequality in Access to Publicly Subsidized Sports Infrastructure: A National Analysis of NFL Stadium Accessibility Across U.S. Counties (2026-08-20)
3 of them
  • How Accessibility Shapes Land Use
  • Measures of Spatial Accessibility to Health Care in a GIS Environment: Synthesis and a Case Study in the Chicago Region
  • An enhanced two-step floating catchment area (E2SFCA) method for measuring spatial accessibility to primary care physicians
; Towards healthy homes: Exploring the effects of housing typologies on mental health in Australian capital cities shares 3 references with From housing prices to depressive symptoms: Serial mediating roles of subjective social status and status anxiety (2026-10-13)
3 of them
  • Housing and Mental Health: A Review of the Evidence and a Methodological and Conceptual Critique
  • Housing Disadvantage and Poor Mental Health: A Systematic Review
  • Housing as a social determinant of health and wellbeing: developing an empirically-informed realist theoretical framework
counted over the 7 papers published above
on the same shoulders
8 of the day's papers cite Travel and the Built Environment (Reid Ewing, Robert Cervero 2010); 4 cite Accessibility evaluation of land-use and transport strategies: review and research directions (Karst Geurs, Bert van Wee 2003); 4 cite Travel demand and the 3Ds: Density, diversity, and design (Robert Cervero, Kara M. Kockelman 1997) counted over the 94 of this day's 120 named papers whose reference lists we have — standing on the same older work is not the same as being about the same thing
first appearance
Primary HealthCare; Technical University of Crete; Xuegang none of them on a paper we published in the 135 days before this issue

Two counts, not one counted twice: every row here but “on the same shoulders” counts the papers published above, and that one counts every paper this day named — those and the ones under Also published today, whose abstracts we could not read.

"Deep Learning-Based Road Lane Quality Assessment from Egocentric Dashcam Video" introduced a baseline framework and a geo-referenced dataset of 63,839 frames from South Texas to automate the evaluation of lane-marking quality. "Demographics-Informed Neural Network for Multimodal Spatiotemporal Forecasting of Urban Growth and Travel Patterns Using Satellite Imagery" presented a deep learning framework that uses past image sequences to predict next-year county-level satellite imagery while maintaining demographic consistency. "Does telehealth narrow the urban-rural gap? A dual assessment of in-person and telehealth prenatal care accessibility" analyzed 2020–2021 Medicaid claims across 424 South Carolina ZIP Code Tabulation Areas, finding that observed prenatal care utilization was significantly lower in rural areas despite no significant urban-rural difference in predicted in-person accessibility.

codedatapreprint

Shaif Chowdhury, Akhila Yeruvasu, Moshiur Rahman Tonmoy, et al.
arXiv cs.CV

deep learning · street view imagery

Deep Learning-Based Road Lane Quality Assessment from Egocentric Dashcam Video

The authors developed a baseline framework and a geo-referenced smartphone dashcam dataset containing 63,839 frames from South Texas to automate lane-marking quality evaluations. The approach introduces three specific quality metrics: Contrast Stability Index ($R^2=0.97$), Line Continuity Ratio ($R^2=0.84$), and Regression Lane Confidence Score based on 3,339 OpenStreetMap-matched frames. In a pilot test of 50 images, these metrics achieved stronger rank correlation with human quality ratings (Spearman's $\rho=0.7316$ for continuity) than standard image-quality algorithms.

Why it matters — Road infrastructure monitoring can shift from manual field inspections to automated, low-cost video surveys using metrics specifically tailored to lane visibility and map consistency rather than general image metrics.

Caveat: Human validation of the proposed quality metrics relies on a small pilot sample of 50 ranked images within a single geographic region.

arXiv · pdf · code

published: Journal of Transportation Engineering Part A Systems

Eugene Kofi Okrah Denteh, Andrews Danyo, Joshua Kofi Asamoah, et al.
Journal of Transportation Engineering Part A Systems · was arXiv

North Dakota State University

Remote Sensing and Land Use · Urban Design and Spatial Analysis · Land Use and Ecosystem Services · deep learning · satellite imagery · United States

Demographics-Informed Neural Network for Multimodal Spatiotemporal Forecasting of Urban Growth and Travel Patterns Using Satellite Imagery

The study introduces the Demographics-Informed Neural Network (DINN), a deep learning framework designed to predict next-year county-level satellite imagery from past image sequences while maintaining demographic consistency. Evaluated on US county data spanning 2012 to 2023, the model achieved a satellite image prediction MSE of 0.00379 and an SSIM of 0.8342, outperforming ConvLSTM, transformer-based, and encoder-decoder baselines. Additionally, its demographic predictor explained roughly 80% of the variance in population characteristics, which were then used to forecast aggregate county-level vehicle availability and commute-mode counts.

Why it matters — It demonstrates that satellite imagery sequence models can joint-predict spatial raster changes and aggregate demographic and transportation indicators, offering a unified forecasting pipeline across US counties.

Caveat: The model's transfer learning performs significantly better on absolute counts than on scale-controlled shares, meaning its predictions largely track total population scale and reflect county-level associations rather than direct causal behavior.

Journal of Transportation Engineering Part A Systems · doi

published: Applied Geography

Dikshya Panta, Sicheng Wang, Yanfeng Xu, et al.
Applied Geography

University of South Carolina

Telemedicine and Telehealth Implementation · Maternal and Neonatal Healthcare · Global Health Workforce Issues · spatial econometrics · University of South Carolina · Primary HealthCare

Does telehealth narrow the urban-rural gap? A dual assessment of in-person and telehealth prenatal care accessibility

Analysis of 2020-2021 Medicaid claims across 424 ZIP Code Tabulation Areas in South Carolina demonstrates that predicted in-person prenatal accessibility showed no significant urban-rural difference, yet observed utilization was significantly lower in rural areas (p < .001). Telehealth accessibility remained heavily skewed toward urban areas, with zero predicted telehealth access recorded in 36.76% of rural areas compared to 7.53% of urban ones. Researchers combined broadband availability constraints with floating catchment area methods and 30-minute visit volumes, assessing spatial associations through spatial lag and error models.

Why it matters — The findings counter the assumption that virtual care automatically overcomes geographic health barriers, demonstrating that telehealth operates within and mirrors existing in-person obstetric deficits rather than evening them out.

Caveat: The study relies on Medicaid claims from a single state during the early pandemic period, which may reflect unique healthcare utilization dynamics and policy waivers.

Applied Geography · doi

published: Cities

Sook Yan Wong, Erika Martino, Ang Li
Cities

The University of Melbourne

Urban Green Space and Health · Urban Transport and Accessibility · Health disparities and outcomes · The University of Melbourne

Towards healthy homes: Exploring the effects of housing typologies on mental health in Australian capital cities

Using fixed-effects regression models on HILDA panel data spanning 2001 to 2019, this study evaluated the joint impact of housing structure and tenancy type on long-term mental health across major Australian metropolitan areas. Across the country, no statistically significant overall difference in mental health was found between apartment dwellers and house residents, or between renters and owners. However, city-specific models revealed that apartment residents in Greater Sydney experienced slightly better mental health outcomes compared to those residing in detached houses.

Why it matters — This finding challenges the widespread assumption that apartment living inherently worsens mental health, demonstrating instead that structure type has no universal direct negative effect and that mental health outcomes depend heavily on local urban context.

Caveat: The findings rely on self-reported survey data and focus on broad population averages, which may obscure intra-city micro-environmental conditions or building-specific quality issues.

Cities · doi

published: Cities

Aikaterini Lilli, Fani Konstantinou, Elisavet Tsekeri, et al.
Cities

Technical University of Crete

Urban Green Space and Health · Urban Transport and Accessibility · Health disparities and outcomes · regression · Technical University of Crete

Beyond proximity: Green space visitation patterns and physical health outcomes

A cross-sectional survey of 718 adult residents in Chania, Crete, conducted from June to August 2024, evaluated the relationship between green space visitation, demographics, and physical health conditions including BMI, coronary artery disease, hypertension, and hypercholesterolemia. Logistic regression models revealed that living more than a 20-minute walk from green space increased the odds of hypertension (OR = 1.603), while visits lasting longer than one hour were linked to lower odds of coronary artery disease, hypertension, and hypercholesterolemia alongside increased physical activity duration.

Why it matters — The study demonstrates that visit duration and specific green space types correlate with metabolic and cardiovascular health outcomes independently of simple spatial proximity in a Mediterranean urban environment.

Caveat: The findings rely on cross-sectional self-reported survey data from a single city, which limits causal inference regarding green space visitation and health outcomes.

Cities · pdf · doi

datapreprint

Eleanor A. Power, Daniel Redhead, Tom Rutter, et al.
arXiv cs.SI

A cross-cultural dataset on the structure of social networks: The ENDOW project

The authors assembled a cross-cultural dataset of multilayer social networks across 50 communities in 29 countries, covering diverse livelihoods including fishers, pastoralists, horticulturalists, farmers, and wage laborers. Household-level surveys recorded ten types of material and immaterial support exchanges to construct inter-unit social networks. Comparing network summary statistics and nomination patterns across these communities and against benchmark datasets reveals high structural heterogeneity.

Why it matters — This dataset provides empirical baseline metrics for non-industrialized and culturally diverse populations, challenging universal assumptions and standard rules of thumb regarding human social network structure.

arXiv · pdf

preprint

Zepu Wang, Luyao Niu, Ruolin Li, et al.
arXiv cs.LG

Xuegang

GeoSplit: Selective Edge--Server Map Acquisition for Cross-User Travel-Mode Recognition

The researchers developed GeoSplit, an edge-server framework that optimizes map data retrieval during travel-mode classification from mobility trajectories. Tested on the GeoLife dataset with 10% labeled data under user-disjoint splitting, full map context boosted macro-F1 from 0.7154 to 0.7864, whereas an entropy-gated query model reached a 0.7729 macro-F1 while requesting map data for only 39.5% of trajectories and reducing network payload by 60.5%.

Why it matters — The system demonstrates that edge devices can maintain high motion classification accuracy while dramatically reducing mobile data transfer, bypassing the need to constantly download map context or continuously send raw trajectory streams to a server.

Caveat: Neither learned value estimation nor probability fusion provided stable performance gains over simpler gating mechanisms across the tested split protocols.

arXiv · pdf

Still cited

Walter G Hansen (1959), How Accessibility Shapes Land Use 1 of today's items cite it · 42 of 3384 in the archive stand on it

Also published today

These appeared today in journals we track. Their abstracts are not openly available, so we cannot summarise them.