Graph Query Networks for Object Detection with Automotive Radar

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Saini, Loveneet, Tercan, Hasan, Meisen, Tobias
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866912718657683456
author Saini, Loveneet
Tercan, Hasan
Meisen, Tobias
author_facet Saini, Loveneet
Tercan, Hasan
Meisen, Tobias
contents Object detection with 3D radar is essential for 360-degree automotive perception, but radar's long wavelengths produce sparse and irregular reflections that challenge traditional grid and sequence-based convolutional and transformer detectors. This paper introduces Graph Query Networks (GQN), an attention-based framework that models objects sensed by radar as graphs, to extract individualized relational and contextual features. GQN employs a novel concept of graph queries to dynamically attend over the bird's-eye view (BEV) space, constructing object-specific graphs processed by two novel modules: EdgeFocus for relational reasoning and DeepContext Pooling for contextual aggregation. On the NuScenes dataset, GQN improves relative mAP by up to +53%, including a +8.2% gain over the strongest prior radar method, while reducing peak graph construction overhead by 80% with moderate FLOPs cost.
format Preprint
id arxiv_https___arxiv_org_abs_2511_15271
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Graph Query Networks for Object Detection with Automotive Radar
Saini, Loveneet
Tercan, Hasan
Meisen, Tobias
Computer Vision and Pattern Recognition
Machine Learning
Object detection with 3D radar is essential for 360-degree automotive perception, but radar's long wavelengths produce sparse and irregular reflections that challenge traditional grid and sequence-based convolutional and transformer detectors. This paper introduces Graph Query Networks (GQN), an attention-based framework that models objects sensed by radar as graphs, to extract individualized relational and contextual features. GQN employs a novel concept of graph queries to dynamically attend over the bird's-eye view (BEV) space, constructing object-specific graphs processed by two novel modules: EdgeFocus for relational reasoning and DeepContext Pooling for contextual aggregation. On the NuScenes dataset, GQN improves relative mAP by up to +53%, including a +8.2% gain over the strongest prior radar method, while reducing peak graph construction overhead by 80% with moderate FLOPs cost.
title Graph Query Networks for Object Detection with Automotive Radar
topic Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2511.15271