Conference Agenda

This contribution introduces GT-LOD3, a new benchmark dataset designed to advance semantic Level of Detail 3 (LOD3) building reconstruction from UAS-based photogrammetric point clouds. Existing benchmarks primarily focus on mesh- or point-level semantic labelling, façade segmentation, or LOD2-level modelling, but high-quality, geometry-accurate LOD3 ground truth paired with real-world photogrammetric observations are still limited. GT-LOD3 fills this gap by offering paired UAS point clouds and manually modeled LOD3 reference data in CityGML format, enabling research on window-level facade reasoning, geometric regularization, and instance-level shape recovery.

The benchmark currently consists of two subsets featuring different architectural styles and environmental conditions: (1) a urban block in Gold Coast (Lakewood, Ohio, USA), and (2) the Technical University of Munich (TUM) campus. The accompanying LOD3 reference models contain explicit window geometry, enabling detailed evaluation of both detection performance and polygon-level geometric accuracy.

We further provide a baseline reconstruction pipeline that combines point-cloud semantic segmentation, facade-aligned 2D projection, window region extraction, and geometric back-projection into CityGML. An evaluation protocol is presented including pixel-level metrics (IoU, precision, recall, F1) and instance-level detection metrics based on optimal assignment via the Hungarian algorithm.

Building wireframe reconstruction from LiDAR faces challenges due to sparse and incomplete point cloud data. We present LoD2-Former, a multi-modal Transformer architecture that fuses aerial LiDAR and optical imagery for end-to-end 3D roof wireframe reconstruction. Unlike existing point-cloud-only methods, our dual-backbone approach with bidirectional cross-modal attention leverages complementary geometric and visual information. Experiments on two datasets show consistent improvements in edge detection metrics, with edge F1-scores increasing from 0.874 to 0.899 on Tallinn and 0.968 to 0.974 on Roof-Intuitive, while substantially boosting corner recall (0.630 to 0.729) in complete-data settings. We also contribute a curated multi-modal subset of Building3D with aligned LiDAR and aerial imagery to facilitate future research.

In this paper, a method exploiting aerial LiDAR point clouds to build realistic building meshes suitable for electromagnetic simulation is proposed. One of the main challenges lies in reconstructing regularized building meshes with low polygonal density. Optimization-based methods, commonly used for building reconstruction from point clouds, are highly data-driven, making the quality of results dependent on the quality of input data. Aerial LiDAR scans can be incomplete or sparse, for instance due to occlusion. A novel LoD2 buildings reconstruction method based on deep learning is proposed, assuming that deep learning methods are more robust to incomplete or sparse data than optimization-based methods. A parametric building model is introduced, based on the Weighted Straight Skeleton algorithm, which generates realistic roofs from a building footprint and an associated set of slopes, and subsequently extrudes the roof to the specified building height. This parametric approach guarantees that a given set of parameters (height, footprint and slopes) produces a regularized building mesh with low polygonal density. A multimodal model, named Point2WSS, was trained to recover the variable number of building's continuous parameters from its corresponding point cloud. This approach enables the generation of realistic building meshes suitable for electromagnetic simulation, if the predicted parameters accurately approximate real-world values.

The modeling of buildings suffers from a dichotomy between generic and specific representations: the lack of domain knowledge in flexible models that can represent many shapes, and the restricted geometry of pre-specified parametric building primitives. To fill this gap, we propose using general boundary representations enriched with automatically recognized and enforced geometric constraints derived from human-made regularities. The proposed reasoning process relies on the statistics of the planar point groups extracted from airborne-captured point clouds. Hence, a chosen significance level is the only process parameter. To enforce the creation of sound solids, we apply manifold constraints for the generation of the boundary representations. The feasibility and usability of the approach are demonstrated by evaluating an airborne-captured laser scan containing approximately 7,600 buildings over an area of 50 km^2 featuring both inner-city and rural landscapes.

We present P3, a large-scale multimodal dataset for building vectorization, including aerial LiDAR point clouds, aerial images, and vectorized 2D building outlines, collected across three continents. P3 contains over 10 billion LiDAR points with decimeter-level accuracy and RGB images at a ground sampling distance of 25 centimeters. While many existing datasets focus on the image modality, P3 offers a complementary perspective by incorporating dense 3D information. We demonstrate that LiDAR point clouds serve as a robust modality for predicting building polygons, both in hybrid and end-to-end learning frameworks. Moreover, fusing LiDAR and imagery further improves accuracy and geometric quality of predicted polygons. The P3 dataset is publicly available, along with code and pretrained weights of three state-of-the-art models for building polygon prediction at https://github.com/raphaelsulzer/PixelsPointsPolygons .

Accurate building height estimation plays a crucial role in large-scale 3D urban reconstruction. However, conventional stereo matching approaches often suffer from mismatches around building edges, leading to unreliable height retrieval in dense urban areas. To address this issue, this paper presents a novel method for building height estimation based on contour vector registration integrated with the vertical line locus technique. The proposed framework first automatically matches building contour vectors extracted from stereo high-resolution satellite images. Then, for each paired contour, a range of candidate heights is searched using a rational function model to project the reference contour from the image space to object space and then reproject it onto the conjugate image. The elevation that maximizes the overlap ratio between projected and paired contours is identified as the optimal roof elevation. Building height is subsequently derived by subtracting the ground elevation from the estimated roof elevation. Experiments conducted on SuperView-1 (SV-1) satellite stereo images over Jiuyuan District, Baotou, Inner Mongolia, China, demonstrate the effectiveness of the proposed method. The resulting building height estimates achieve a root mean square error of 0.84 m compared to manual measurements, showing strong agreement (r = 0.9993). The proposed contour-based stereo registration approach provides a robust and efficient solution for building height extraction from high-resolution satellite data, supporting precise urban 3D modeling and large-scale spatial analysis.

Urban development in the Greater Toronto Area faces significant challenges because zoning abbreviations and terminology vary widely between municipalities. This provides the background and motivation for the study, as labels such as “R2” in Toronto and “R2 S” in Markham appear similar yet represent different permissions and development standards, creating confusion and slowing planning workflows in a region with growing housing pressures. The problem addressed in this research is the absence of a unified, machine-readable framework that standardizes zoning terminology across municipalities, which limits automated compliance checking, GIS integration, and cross municipal comparison. The objective of this work is to create Zonology, an ontology-based framework that harmonizes zoning abbreviations, permitted land uses, and development standards, beginning with the City of Toronto and the City of Markham. The methodology follows the Linked Open Terms approach, using the Web Ontology Language to encode zoning by laws, land use categories, development standards, and spatial relationships. The model is evaluated through reasoning tasks, competency questions, and semantic alignment tests to ensure clarity, consistency, and interoperability. The results show that Zonology successfully aligns more than sixty zoning categories and over one hundred fifty land use permissions, enabling consistent semantic interpretation and cross municipal queries. The overall significance of this work is that the ontology improves regulatory clarity, strengthens data driven planning, and provides a scalable foundation for harmonized zoning governance across the Greater Toronto Area.

Urban systems are increasingly complex, interconnected, and dynamic, yet most geospatial data models continue to represent them as static geometric layers with limited support for explicit relationships and semantics. This restricts advanced spatial reasoning, network analysis, and AI-driven applications.

This paper introduces GeoGraphJSON, a lightweight semantic data model that extends GeoJSON by integrating spatial geometry with graph-based connectivity. The framework represents spatial entities as nodes and explicitly encodes relationships as typed edges, enabling unified representation of geometry, topology, and semantics within a single interoperable structure. A hierarchical Unique Identifier (UID) system ensures consistent lineage and cross-layer integration across administrative, transportation, and urban asset datasets.

The approach is validated using a large-scale urban dataset from Riyadh, comprising over 10,000 nodes and 13,000 edges. Graph-based analysis demonstrates realistic spatial patterns, including right-skewed degree distribution, strong network connectivity, and identifiable community structures. These results highlight the ability of GeoGraphJSON to capture hierarchical organization and functional relationships while supporting efficient analytical workflows.

By bridging geometry-centric GIS models and graph-based approaches, GeoGraphJSON provides a scalable foundation for urban analytics, digital twins, and GeoAI applications, enabling geospatial systems to evolve from static representations toward intelligent, relationship-aware spatial models.

Urban morphology quantitatively reveals how distinct historical and functional drivers shape city form. This study employs a computational morphometric approach using the Momepy library to analyze and compare the urban structures of Cairo, Egypt, and Makkah, Saudi Arabia. These cities represent paradigmatic cases: Cairo exemplifies long-term, organic layering, while Makkah demonstrates rapid, purpose-driven transformation for religious pilgrimage. We calculated key metrics—including tessellation area, convexity, elongation, equivalent rectangular index, and edge betweenness centrality—for building footprints and street networks sourced from OpenStreetMap. Results show Cairo possesses a heterogeneous, polycentric fabric with complex plot shapes and a distributed street network, reflecting its layered history. Conversely, Makkah exhibits a more monocentric, consolidated form with standardized building geometries and a hierarchical street network channeling movement toward its core. The findings demonstrate that quantitative morphology effectively captures how Cairo's organic evolution and Makkah's centralized planning produce fundamentally different, yet equally revealing, urban spatial structures, offering a replicable framework for cross-city analysis in the region

Urban illumination, as captured through Nighttime Light (NTL) data, serves as a powerful indicator of human activity, infrastructure development, and socioeconomic progress in rapidly growing cities. However, previous studies on Delhi have largely focused on temporal NTL trends without integrating multi-year statistical and spatial analyses to reveal underlying urban and socioeconomic dynamics. This study investigates the spatiotemporal dynamics of urban illumination and development over Delhi using VIIRS Day/Night Band (DNB) NTL data for the years 2015, 2020, and 2025. NTL intensity was used as a proxy for urbanization and socioeconomic activity. Monthly composite datasets for January of each year were processed, clipped to the Delhi administrative boundary, and analyzed using statistical, temporal, and correlation-based methods. The results revealed a slight decline in mean NTL intensity from 26.34 in 2015 to 24.95 in 2025, indicating stabilization in overall light emissions may be due to the adoption of energy-efficient technologies. However, the maximum and range values increased markedly (166.85 to 228.04), signifying intensified illumination in high-activity commercial and infrastructural zones. Temporal change analysis showed balanced positive and negative illumination shifts, with over 50% of pixels exhibiting moderate growth during 2020–2025. Strong Pearson and Spearman correlations (r = 0.83–0.92; ρ = 0.910.95) confirmed the temporal consistency of illumination distribution. The socioeconomic assessment highlighted spatial disparities in light intensity might be corresponding to varying economic activity levels. Overall, the study demonstrates that VIIRS-derived NTL data provide an effective and robust approach for monitoring urban growth, socioeconomic variability, and sustainable lighting transitions in metropolitan environments.

The research investigates artificial intelligence as a device for the automatic interpretation of landscape, reframing representation not as a neutral reproduction but as a cognitive operation in which perception, description, and evaluation converge. Moving from the assumption that landscape is not an objective given but a culturally and perceptually constructed form, the study proposes a fully data-driven methodology based on geolocated images. Through a systematic grid sampling, street-level panoramic views are collected and processed within a multimodal pipeline integrating visual analysis, language models, and multi-agent evaluation. Images are first translated into textual descriptions and semantically clustered, allowing territorial classes to emerge from the data rather than from predefined taxonomies. In parallel, a simulated cognitive framework, structured through four agent profiles, produces evaluative scores and textual judgments, later analysed through sentiment detection. The integration of these layers generates a georeferenced dataset from which a perceptual cartography of the territory is constructed. Applied to the urban context of San Giustino (Italy), the method reveals a continuous gradient from dense urban cores to rural landscapes, while exposing differentiated perceptual readings across observer profiles. Within this framework, artificial intelligence does not replace human interpretation; it operates as an epistemic extension, transforming the landscape into a distributed field of comparable perceptions, where representation becomes a computable form of shared knowledge.

This study focuses on bridging usability information between data providers and data users through standardized metadata. By further integrating standardized metadata with geographic information system operation design, the operations gain automated and awareness capabilities, allowing usability information based on data specifications and quality considerations to be incorporated into relevant processes, thereby avoiding erroneous decisions. The research references international standards such as ISO 19115 and ISO 19157 to meet the requirements of open geographic information technologies.

Large-scale 3D reconstruction has emerged as a key research in the fields of photogrammetry and computer vision. 3D Gaussian Splatting (3DGS) has become a mainstream approach due to its efficient rendering, but it confronts critical challenges in large-scale scenarios: excessive memory overhead and inadequate geometric accuracy. Meanwhile, the traditional Structure from Motion and Multi-view Stereo (SfM-MVS) framework, despite its cumbersome process, continues to exhibit robust performance. Notably, a systematic evaluation comparing these two paradigms in large-scale scenes remains absent. To address this, we develop a unified verification framework to evaluate the texture rendering quality and geometric reconstruction precision of several recent methods using real-world datasets. The results indicate that SfM-MVS methods still maintain an advantage in the completeness and accuracy of geometric reconstruction. In contrast, 3DGS methods have achieved breakthroughs in local accuracy or rendering-geometry synergy, yet their global consistency requires further improvement.

3DGS-based methods often render transient distractors in 3D scenes as significant floating artifacts. Existing works for removing transient distractors suffer from under-identification or over-identification, resulting in residual transient distractors affecting reconstruction quality or loss of scene information, preventing the reconstruction of fine details. To address these challenges, we propose RobustGauss. We first rely solely on the cosine similarity of DINOv2 features to robustly predict uncertainty masks and accurately identify the main regions of transient disturbances and their corresponding shadows. Due to the limited resolution of DINOv2 features, we use high-resolution image residuals to refine the edges of the initial uncertainty masks, thereby accurately identifying all transient distractors and minimizing their impact on 3D scene reconstruction. Experiments on two challenging datasets demonstrate that our method achieves state-of-the-art performance.

N/A

We present EMVSNet, a sampling-free Multi-View Stereo (MVS) method that, to the best of our knowledge, is the first to integrate Evidential Deep Learning into MVS. Given a set of overlapping images, our method predicts a depth value together with its associated uncertainty per pixel of a reference image, incorporating uncertainty from aleatoric and epistemic sources. Specifically, we use an existing convolutional neural network architecture designed for MVS as backbone and extend it to regress evidential parameters per pixel, describing the probability distribution over the depth corresponding to this pixel. In contrast to existing MVS methods that often neglect epistemic uncertainty or obtain it via sampling at inference, our evidential formulation does not require sampling, but enables single-pass inference. We evaluate the uncertainty estimation capabilities of our method using two publicly available datasets and compare the depth predictions against a deterministic variant. The experimental results demonstrate that EMVSNet achieves competitive depth accuracy while, at the same time, providing uncertainty estimates that enable us to reliably rank depth estimates according to their risk of being incorrect and to automatically identify out of distribution data. Our model shows only slightly increased inference time compared to a deterministic baseline while giving comparable uncertainty estimates to an computationally expensive sampling based approach, marking a first step towards real-time capable uncertainty estimation for image-based 3D reconstruction.

Object completion networks typically produce static Signed Distance Fields (SDFs) that faithfully reconstruct geometry but cannot be rescaled or deformed without introducing structural distortions. This limitation restricts their use in applications requiring flexible object manipulation, such as indoor redesign, simulation, and digital content creation. We introduce a part-aware scaling framework that transforms these static completed SDFs into editable, structurally coherent objects. Starting from SDFs and Texture Fields generated by state-of-the-art completion models, our method performs automatic part segmentation, defines user-controlled scaling zones, and applies smooth interpolation of SDFs, color, and part indices to enable proportional and artifact-free deformation. We further incorporate a repetition-based strategy to handle large-scale deformations while preserving repeating geometric patterns. Experiments on Matterport3D and ShapeNet objects show that our method overcomes the inherent rigidity of completed SDFs and is visually more appealing than global and naive selective scaling, particularly for complex shapes and repetitive structures.

Panoptic segmentation requires the simultaneous recognition of countable thing instances and amorphous stuff regions, placing joint demands on long-range context modelling, multi-scale feature representation, and efficient dense prediction. Existing convolutional and transformer-based methods struggle to satisfy all three requirements concurrently: convolutional architectures are limited in their capacity to model long-range dependencies, while transformer-based methods incur quadratic computational cost that is prohibitive at high resolutions. In this paper, we propose MambaPanoptic, a fully Mamba-based panoptic segmentation framework that addresses these limitations through two principal contributions. First, we introduce MambaFPN, a top-down feature pyramid that leverages Mamba blocks to generate globally coherent, multi-scale feature representations with linear computational complexity. Second, we adopt a PanopticFCN-style kernel generator that produces unified thing and stuff kernels for proposal-free panoptic prediction, enhanced by a QuadMamba-based feature refinement module applied at multiple network stages. Experiments on the Cityscapes and COCO panoptic segmentation benchmarks demonstrate that MambaPanoptic consistently outperforms PanopticDeepLab and PanopticFCN under comparable model sizes, and matches or surpasses Mask2Former on Cityscapes in PQ and AP while requiring fewer parameters.

Single-view 3D reconstruction from monocular aerial imagery presents a fundamental challenge in remote sensing due to the inherent scale ambiguity and the complex geometry of urban environments. Traditional regression-based methods often struggle to recover high-frequency structural details, leading to over-smoothed or noisy outputs. To address this, we introduce GeoPrior-Diff, a novel two-stage framework that leverages the generative capabilities of Latent Diffusion Models to reconstruct high-fidelity 3D point clouds.

Unlike direct generation approaches, our method explicitly bridges the domain gap between 2D texture and 3D structure by utilizing an intermediate geometric prior. In the first stage, we predict an oblique normal map from the input satellite imagery, capturing essential surface orientation and structural boundaries. In the second stage, this normal map serves as a strong conditioning signal for a probabilistic diffusion model, guiding the denoising process to synthesize accurate 3D point clouds. Preliminary results demonstrate that decoupling geometric estimation from point generation significantly enhances structural consistency and reduces artifacts compared to baseline methods. This work highlights the potential of using generative priors for robust 3D urban modeling from limited data.