arXiv:2601.08375v2 Announce Type: replace 
Abstract: Semantic segmentation of 3D geospatial point clouds is fundamental to remote sensing applications, yet domain shifts caused by regional and acquisition-related variations often degrade model performance. Although domain adaptation can mitigate such shifts, existing methods typically require access to source-domain data, which is often infeasible due to privacy concerns and regulatory policies. To address this, we propose LoGo (Local-Global Dual-Consensus), a novel source-free unsupervised domain adaptation (SFUDA) framework requiring only a pretrained model and unlabeled target data. At the local level, we introduce a class-balanced prototype estimation module that ensures that robust feature prototypes can be generated even for sample-scarce tail classes, effectively mitigating the feature collapse caused by long-tailed distributions. At the global level, we introduce an optimal transport-based global distribution alignment module that formulates pseudo-label assignment as a global optimization problem, effectively correcting the over-dominance of head classes inherent in local greedy assignments, and thereby preventing model predictions from being severely biased towards majority classes. Finally, we propose a dual-consistency pseudo-label filtering mechanism that retains only high-confidence pseudo-labels where local multi-augmented ensemble predictions align with global optimal transport assignments for self-training. Extensive experiments on two challenging benchmarks, encompassing cross-scene and cross-sensor settings, demonstrate that LoGo consistently outperforms existing state-of-the-art methods. The source code is available at https://github.com/GYproject/LoGo-SFUDA.

تم اقتراح إطار جديد للتكيف مع المجال بدون مصدر يسمى LoGo لعملية تقسيم السحب النقطية الجغرافية ثلاثية الأبعاد، حيث يعالج التحديات الناتجة عن التحولات في المجال بسبب التغيرات الإقليمية والمتعلقة بالاستحواذ. يتطلب هذا الإطار فقط نموذجًا مدربًا مسبقًا وبيانات مستهدفة غير مصنفة، مما يمثل تقدمًا كبيرًا في تطبيقات الاستشعار عن بعد.

Se ha propuesto un nuevo marco de adaptación de dominio sin fuente llamado LoGo para la segmentación semántica de nubes de puntos geoespaciales 3D, abordando los desafíos que presentan los cambios de dominio debido a variaciones regionales y relacionadas con la adquisición. Este marco requiere solo un modelo preentrenado y datos de destino no etiquetados, lo que representa un avance significativo en las aplicaciones de teledetección.

Un nouveau cadre d'adaptation de domaine sans source, nommé LoGo, a été proposé pour la segmentation sémantique des nuages de points géospatiaux 3D, répondant aux défis posés par les décalages de domaine dus aux variations régionales et liées à l'acquisition. Ce cadre nécessite uniquement un modèle pré-entraîné et des données cibles non étiquetées, ce qui constitue une avancée significative dans les applications de télédétection.

A novel source-free unsupervised domain adaptation framework named LoGo has been proposed for semantic segmentation of 3D geospatial point clouds, addressing the challenges posed by domain shifts due to regional and acquisition-related variations. This framework requires only a pretrained model and unlabeled target data, making it a significant advancement in remote sensing applications.