arXiv:2511.17322v1 Announce Type: new 
Abstract: In this paper, we introduce NoPe-NeRF++, a novel local-to-global optimization algorithm for training Neural Radiance Fields (NeRF) without requiring pose priors. Existing methods, particularly NoPe-NeRF, which focus solely on the local relationships within images, often struggle to recover accurate camera poses in complex scenarios. To overcome the challenges, our approach begins with a relative pose initialization with explicit feature matching, followed by a local joint optimization to enhance the pose estimation for training a more robust NeRF representation. This method significantly improves the quality of initial poses. Additionally, we introduce global optimization phase that incorporates geometric consistency constraints through bundle adjustment, which integrates feature trajectories to further refine poses and collectively boost the quality of NeRF. Notably, our method is the first work that seamlessly combines the local and global cues with NeRF, and outperforms state-of-the-art methods in both pose estimation accuracy and novel view synthesis. Extensive evaluations on benchmark datasets demonstrate our superior performance and robustness, even in challenging scenes, thus validating our design choices.

تم تقديم خوارزمية تحسين جديدة تُدعى NoPe-NeRF++ لتدريب حقول الإشعاع العصبي (NeRF) دون الحاجة إلى أولويات الوضع. تُحسن هذه الطريقة استعادة وضع الكاميرا في السيناريوهات المعقدة من خلال البدء بت初始化 وضع نسبي واستخدام تحسين مشترك محلي، تليه مرحلة تحسين عالمية تتضمن قيود التناسق الهندسي من خلال ضبط الحزمة.

Se ha introducido un nuevo algoritmo de optimización llamado NoPe-NeRF++ para el entrenamiento de Campos de Radiancia Neurales (NeRF) sin necesidad de priors de pose. Este método mejora la recuperación de poses de cámara en escenarios complejos comenzando con una inicialización de pose relativa y empleando una optimización conjunta local, seguida de una fase de optimización global que integra restricciones de consistencia geométrica a través del ajuste de haz.

Un nouvel algorithme d'optimisation nommé NoPe-NeRF++ a été introduit pour l'entraînement des Neural Radiance Fields (NeRF) sans nécessiter de prioris de pose. Cette méthode améliore la récupération des poses de caméra dans des scénarios complexes en commençant par une initialisation de pose relative et en utilisant une optimisation conjointe locale, suivie d'une phase d'optimisation globale qui intègre des contraintes de cohérence géométrique par ajustement en faisceau.

A new optimization algorithm named NoPe-NeRF++ has been introduced for training Neural Radiance Fields (NeRF) without the need for pose priors. This method enhances camera pose recovery in complex scenarios by starting with relative pose initialization and employing local joint optimization, followed by a global optimization phase that integrates geometric consistency constraints through bundle adjustment.

NoPe-NeRF++: Local-to-Global Optimization of NeRF with No Pose Prior

arXiv:2502.05708v2 Announce Type: replace-cross 
Abstract: We present GRaF, Generalizable Radio-Frequency (RF) Radiance Fields, a framework that models RF signal propagation to synthesize spatial spectra at arbitrary transmitter or receiver locations, where each spectrum measures signal power across all surrounding directions at the receiver. Unlike state-of-the-art methods that adapt vanilla Neural Radiance Fields (NeRF) to the RF domain with scene-specific training, GRaF generalizes across scenes to synthesize spectra. To enable this, we prove an interpolation theory in the RF domain: the spatial spectrum from a transmitter can be approximated using spectra from geographically proximate transmitters. Building on this theory, GRaF comprises two components: (i) a geometry-aware Transformer encoder that captures spatial correlations from neighboring transmitters to learn a scene-independent latent RF radiance field, and (ii) a neural ray tracing algorithm that estimates spectrum reception at the receiver. Experimental results demonstrate that GRaF outperforms existing methods on single-scene benchmarks and achieves state-of-the-art performance on unseen scene layouts.

يمثل تقديم مجالات الإشعاع الراديوي القابلة للتعميم، أو GRaF، تقدمًا كبيرًا في نمذجة انتشار إشارات الراديو، مما يسمح بتوليف الأطياف المكانية في مواقع عشوائية للمرسلين أو المستلمين. يستخدم هذا الإطار نظرية استيفاء تقرب الطيف المكاني من المرسلين القريبين، مما يعزز فهم سلوك إشارات الراديو في بيئات متنوعة.

La introducción de los Campos de Radiancia de Radiofrecuencia Generalizables, o GRaF, representa un avance significativo en la modelización de la propagación de señales RF, permitiendo la síntesis de espectros espaciales en ubicaciones arbitrarias de transmisores o receptores. Este marco utiliza una teoría de interpolación que aproxima el espectro espacial a partir de transmisores cercanos, mejorando así la comprensión del comportamiento de las señales RF en diversos entornos.

L'introduction des Champs de Radiance Radio-Fréquence Généralisables, ou GRaF, représente une avancée significative dans la modélisation de la propagation des signaux RF, permettant la synthèse de spectres spatiaux à des emplacements arbitraires d'émetteurs ou de récepteurs. Ce cadre utilise une théorie d'interpolation qui approxime le spectre spatial à partir d'émetteurs voisins, améliorant ainsi la compréhension du comportement des signaux RF dans divers environnements.

The introduction of Generalizable Radio-Frequency (RF) Radiance Fields, or GRaF, marks a significant advancement in modeling RF signal propagation, allowing for the synthesis of spatial spectra at arbitrary transmitter or receiver locations. This framework utilizes an interpolation theory that approximates the spatial spectrum from nearby transmitters, enhancing the understanding of RF signal behavior in various environments.

NoPe-NeRF++: Local-to-Global Optimization of NeRF with No Pose Prior

Was this article worth reading? Share it