TacoDepth: Towards Efficient Radar-Camera Depth Estimation with One-stage Fusion

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Wang, Yiran, Li, Jiaqi, Hong, Chaoyi, Li, Ruibo, Sun, Liusheng, Song, Xiao, Wang, Zhe, Cao, Zhiguo, Lin, Guosheng
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866917986726576128
author Wang, Yiran
Li, Jiaqi
Hong, Chaoyi
Li, Ruibo
Sun, Liusheng
Song, Xiao
Wang, Zhe
Cao, Zhiguo
Lin, Guosheng
author_facet Wang, Yiran
Li, Jiaqi
Hong, Chaoyi
Li, Ruibo
Sun, Liusheng
Song, Xiao
Wang, Zhe
Cao, Zhiguo
Lin, Guosheng
contents Radar-Camera depth estimation aims to predict dense and accurate metric depth by fusing input images and Radar data. Model efficiency is crucial for this task in pursuit of real-time processing on autonomous vehicles and robotic platforms. However, due to the sparsity of Radar returns, the prevailing methods adopt multi-stage frameworks with intermediate quasi-dense depth, which are time-consuming and not robust. To address these challenges, we propose TacoDepth, an efficient and accurate Radar-Camera depth estimation model with one-stage fusion. Specifically, the graph-based Radar structure extractor and the pyramid-based Radar fusion module are designed to capture and integrate the graph structures of Radar point clouds, delivering superior model efficiency and robustness without relying on the intermediate depth results. Moreover, TacoDepth can be flexible for different inference modes, providing a better balance of speed and accuracy. Extensive experiments are conducted to demonstrate the efficacy of our method. Compared with the previous state-of-the-art approach, TacoDepth improves depth accuracy and processing speed by 12.8% and 91.8%. Our work provides a new perspective on efficient Radar-Camera depth estimation.
format Preprint
id arxiv_https___arxiv_org_abs_2504_11773
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle TacoDepth: Towards Efficient Radar-Camera Depth Estimation with One-stage Fusion
Wang, Yiran
Li, Jiaqi
Hong, Chaoyi
Li, Ruibo
Sun, Liusheng
Song, Xiao
Wang, Zhe
Cao, Zhiguo
Lin, Guosheng
Computer Vision and Pattern Recognition
Radar-Camera depth estimation aims to predict dense and accurate metric depth by fusing input images and Radar data. Model efficiency is crucial for this task in pursuit of real-time processing on autonomous vehicles and robotic platforms. However, due to the sparsity of Radar returns, the prevailing methods adopt multi-stage frameworks with intermediate quasi-dense depth, which are time-consuming and not robust. To address these challenges, we propose TacoDepth, an efficient and accurate Radar-Camera depth estimation model with one-stage fusion. Specifically, the graph-based Radar structure extractor and the pyramid-based Radar fusion module are designed to capture and integrate the graph structures of Radar point clouds, delivering superior model efficiency and robustness without relying on the intermediate depth results. Moreover, TacoDepth can be flexible for different inference modes, providing a better balance of speed and accuracy. Extensive experiments are conducted to demonstrate the efficacy of our method. Compared with the previous state-of-the-art approach, TacoDepth improves depth accuracy and processing speed by 12.8% and 91.8%. Our work provides a new perspective on efficient Radar-Camera depth estimation.
title TacoDepth: Towards Efficient Radar-Camera Depth Estimation with One-stage Fusion
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2504.11773