This study investigates the classification of rooftop shapes using convolutional neural networks (CNNs), with a particular focus on regional adaptation through transfer learning. Initial training of the AlexNet architecture utilizes a publicly available Zenodo dataset comprising satellite imagery of flat, gabled, and hipped roofs. To address generalizability constraints in specific geographic contexts, a custom dataset of Uzbekistan rooftops, sourced from OpenStreetMap and high-resolution Mapbox Static Images API tiles, enables fine-tuning. Experimental outcomes reveal enhanced classification accuracy for flat, gabled, and hipped roofs, underscoring the efficacy of transfer learning in mitigating domain shift due to architectural and environmental variations. Integration of open-source geospatial tools with transfer learning offers a replicable framework for addressing geographic bias in rooftop shape classification, adaptable to other underrepresented regions.
| Mualliflar | Yuldashev, S.U., Юлдашев, С.У. |
|---|---|
| Jurnal | Ҳисоблаш ва амалий математика муаммолари |
| Nashr sanasi | 2025-07-27 |
| Son | 3 |
| Betlar | 133-146 |
| Til | Ingliz |
| DOI | 10.71310/pcam.3_67.2025.12 |
DOI: 10.71310/pcam.3_67.2025.12 · Maqolaning asl sahifasi
roof shape classification, convolutional neural networks, transfer learning, AlexNet, satellite imagery, классификация форм крыш, сверточные нейронные сети, трансферное обучение, AlexNet, спутниковые снимки
Usually, the solution of a plane problem of the theory of elasticity in stresses is reduced to solving a biharmonic equation for the Airy stress function. In this paper, two (A and B) variants of plane boundary value…
Outlier detection is considered in a one-class classification problem. The challenges of detection are related to the choice of loss functions, optimization criteria, and transformations of various types of features…
This article presents a comprehensive mathematical modeling of thermo-electro magneto-elastic processes in anisotropic thin plates of complex shapes using the R function method (RFM). The model considers interrelated…
This study explores the optimization of quadrature formulas for the approximate integration of periodic functions within a specific functional space. The research focuses on developing optimal quadrature formulas by…
Many applied problems are described by the Dirichlet problem for the Poisson equa tions. This problem is used in the following practical situations: membrane or plate deflection, electric potential, gravitational and…
This study is devoted to the development of a system of equations for determining the coefficients of optimal interpolation formulas that integrate derivative information within a Hilbert space framework. Conventional…
This work proposes an expanded 3D Euler model for calculating the spatiotemporal evolution of pollutant concentrations released into the atmosphere under changing meteorological conditions. The model, alongside the…
There are algebraic and variational approaches of construction in the spline theory. In algebraic approach splines are considered as some smooth piecewise polynomial functions. In the variational approach splines are…
This article presents an integrated mathematical model of radial filtration in a cylindrical filter, in which the flow movement is carried out from the outer layer to the inner cavity. Free flow areas are described by…
This study addresses a mixed problem for a quasilinear system of hyperbolic equations expressed in Riemann invariants, incorporating dissipative nonlinear boundary conditions. A numerical approach is developed through…