SecondPose: SE(3)-Consistent Dual-Stream Feature Fusion for Category-Level Pose Estimation

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Chen, Yamei, Di, Yan, Zhai, Guangyao, Manhardt, Fabian, Zhang, Chenyangguang, Zhang, Ruida, Tombari, Federico, Navab, Nassir, Busam, Benjamin
Natura: Preprint
Pubblicazione: 2023
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866910378334617600
author Chen, Yamei
Di, Yan
Zhai, Guangyao
Manhardt, Fabian
Zhang, Chenyangguang
Zhang, Ruida
Tombari, Federico
Navab, Nassir
Busam, Benjamin
author_facet Chen, Yamei
Di, Yan
Zhai, Guangyao
Manhardt, Fabian
Zhang, Chenyangguang
Zhang, Ruida
Tombari, Federico
Navab, Nassir
Busam, Benjamin
contents Category-level object pose estimation, aiming to predict the 6D pose and 3D size of objects from known categories, typically struggles with large intra-class shape variation. Existing works utilizing mean shapes often fall short of capturing this variation. To address this issue, we present SecondPose, a novel approach integrating object-specific geometric features with semantic category priors from DINOv2. Leveraging the advantage of DINOv2 in providing SE(3)-consistent semantic features, we hierarchically extract two types of SE(3)-invariant geometric features to further encapsulate local-to-global object-specific information. These geometric features are then point-aligned with DINOv2 features to establish a consistent object representation under SE(3) transformations, facilitating the mapping from camera space to the pre-defined canonical space, thus further enhancing pose estimation. Extensive experiments on NOCS-REAL275 demonstrate that SecondPose achieves a 12.4% leap forward over the state-of-the-art. Moreover, on a more complex dataset HouseCat6D which provides photometrically challenging objects, SecondPose still surpasses other competitors by a large margin.
format Preprint
id arxiv_https___arxiv_org_abs_2311_11125
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle SecondPose: SE(3)-Consistent Dual-Stream Feature Fusion for Category-Level Pose Estimation
Chen, Yamei
Di, Yan
Zhai, Guangyao
Manhardt, Fabian
Zhang, Chenyangguang
Zhang, Ruida
Tombari, Federico
Navab, Nassir
Busam, Benjamin
Computer Vision and Pattern Recognition
Category-level object pose estimation, aiming to predict the 6D pose and 3D size of objects from known categories, typically struggles with large intra-class shape variation. Existing works utilizing mean shapes often fall short of capturing this variation. To address this issue, we present SecondPose, a novel approach integrating object-specific geometric features with semantic category priors from DINOv2. Leveraging the advantage of DINOv2 in providing SE(3)-consistent semantic features, we hierarchically extract two types of SE(3)-invariant geometric features to further encapsulate local-to-global object-specific information. These geometric features are then point-aligned with DINOv2 features to establish a consistent object representation under SE(3) transformations, facilitating the mapping from camera space to the pre-defined canonical space, thus further enhancing pose estimation. Extensive experiments on NOCS-REAL275 demonstrate that SecondPose achieves a 12.4% leap forward over the state-of-the-art. Moreover, on a more complex dataset HouseCat6D which provides photometrically challenging objects, SecondPose still surpasses other competitors by a large margin.
title SecondPose: SE(3)-Consistent Dual-Stream Feature Fusion for Category-Level Pose Estimation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2311.11125