Salvato in:
Dettagli Bibliografici
Autori principali: Di, Yan, Zhang, Chenyangguang, Wang, Chaowei, Zhang, Ruida, Zhai, Guangyao, Li, Yanyan, Fu, Bowen, Ji, Xiangyang, Gao, Shan
Natura: Preprint
Pubblicazione: 2023
Soggetti:
Accesso online:https://arxiv.org/abs/2311.11106
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866910360581177344
author Di, Yan
Zhang, Chenyangguang
Wang, Chaowei
Zhang, Ruida
Zhai, Guangyao
Li, Yanyan
Fu, Bowen
Ji, Xiangyang
Gao, Shan
author_facet Di, Yan
Zhang, Chenyangguang
Wang, Chaowei
Zhang, Ruida
Zhai, Guangyao
Li, Yanyan
Fu, Bowen
Ji, Xiangyang
Gao, Shan
contents In this paper, we present ShapeMatcher, a unified self-supervised learning framework for joint shape canonicalization, segmentation, retrieval and deformation. Given a partially-observed object in an arbitrary pose, we first canonicalize the object by extracting point-wise affine-invariant features, disentangling inherent structure of the object with its pose and size. These learned features are then leveraged to predict semantically consistent part segmentation and corresponding part centers. Next, our lightweight retrieval module aggregates the features within each part as its retrieval token and compare all the tokens with source shapes from a pre-established database to identify the most geometrically similar shape. Finally, we deform the retrieved shape in the deformation module to tightly fit the input object by harnessing part center guided neural cage deformation. The key insight of ShapeMaker is the simultaneous training of the four highly-associated processes: canonicalization, segmentation, retrieval, and deformation, leveraging cross-task consistency losses for mutual supervision. Extensive experiments on synthetic datasets PartNet, ComplementMe, and real-world dataset Scan2CAD demonstrate that ShapeMaker surpasses competitors by a large margin.
format Preprint
id arxiv_https___arxiv_org_abs_2311_11106
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle ShapeMatcher: Self-Supervised Joint Shape Canonicalization, Segmentation, Retrieval and Deformation
Di, Yan
Zhang, Chenyangguang
Wang, Chaowei
Zhang, Ruida
Zhai, Guangyao
Li, Yanyan
Fu, Bowen
Ji, Xiangyang
Gao, Shan
Computer Vision and Pattern Recognition
In this paper, we present ShapeMatcher, a unified self-supervised learning framework for joint shape canonicalization, segmentation, retrieval and deformation. Given a partially-observed object in an arbitrary pose, we first canonicalize the object by extracting point-wise affine-invariant features, disentangling inherent structure of the object with its pose and size. These learned features are then leveraged to predict semantically consistent part segmentation and corresponding part centers. Next, our lightweight retrieval module aggregates the features within each part as its retrieval token and compare all the tokens with source shapes from a pre-established database to identify the most geometrically similar shape. Finally, we deform the retrieved shape in the deformation module to tightly fit the input object by harnessing part center guided neural cage deformation. The key insight of ShapeMaker is the simultaneous training of the four highly-associated processes: canonicalization, segmentation, retrieval, and deformation, leveraging cross-task consistency losses for mutual supervision. Extensive experiments on synthetic datasets PartNet, ComplementMe, and real-world dataset Scan2CAD demonstrate that ShapeMaker surpasses competitors by a large margin.
title ShapeMatcher: Self-Supervised Joint Shape Canonicalization, Segmentation, Retrieval and Deformation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2311.11106