HOI4D: A 4D Egocentric Dataset for Category-Level Human-Object Interaction

Liu, Yunze; Liu, Yun; Che, Jing; Lyu, Kangbo; Wan, Wen‐Ming; Shen, Hao; Liang, Boqiang; Fu, Zhoujie; Wang, He; Li, Yang

doi:10.48550/arxiv.2203.01577

Cited by 1 publication

(2 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The first five datasets are selected according to previous works on articulated object pose estimation or part decomposition Li et al (2020a); Kawana et al (2021). To further test the effectiveness of our method on objects collected from the real world, we choose two more categories (Safe and Laptop (R)) from a real dataset Liu et al (2022b).…”

Section: Dfs Order Kinematic Chainmentioning

confidence: 99%

See 1 more Smart Citation

Self-Supervised Category-Level Articulated Object Pose Estimation with Part-Level SE(3) Equivariance

Liu¹,

Zhang²,

Hu³

et al. 2023

Preprint

View full text Add to dashboard Cite

Category-level articulated object pose estimation aims to estimate a hierarchy of articulation-aware object poses of an unseen articulated object from a known category. To reduce the heavy annotations needed for supervised learning methods, we present a novel self-supervised strategy that solves this problem without any human labels. Our key idea is to factorize canonical shapes and articulated object poses from input articulated shapes through part-level equivariant shape analysis. Specifically, we first introduce the concept of part-level SE(3) equivariance and devise a network to learn features of such property. Then, through a carefully designed fine-grained pose-shape disentanglement strategy, we expect that canonical spaces to support pose estimation could be induced automatically. Thus, we could further predict articulated object poses as per-part rigid transformations describing how parts transform from their canonical part spaces to the camera space. Extensive experiments demonstrate the effectiveness of our method on both complete and partial point clouds from synthetic and real articulated object datasets. The project page with code and more information can be found at: equi-articulated-pose.github.io.

show abstract

Section: Dfs Order Kinematic Chainmentioning

confidence: 99%

“…We choose seven categories from three different datasets, namely Oven, Washing Machine, Eyeglasses, Laptop (S) with revolute parts from Shape2MotionWang et al (2019b), Drawer with prismatic parts fromSAPIEN Xiang et al (2020), Safe and Laptop (R) with revolute parts from HOI4DLiu et al (2022b).…”

mentioning

confidence: 99%