Saved in:
Bibliographic Details
Main Authors: Guo, Mengqi, Li, Chen, Zhao, Yuyang, Lee, Gim Hee
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2407.15648
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914881366654976
author Guo, Mengqi
Li, Chen
Zhao, Yuyang
Lee, Gim Hee
author_facet Guo, Mengqi
Li, Chen
Zhao, Yuyang
Lee, Gim Hee
contents Inferring step-wise actions to assemble 3D objects with primitive bricks from images is a challenging task due to complex constraints and the vast number of possible combinations. Recent studies have demonstrated promising results on sequential LEGO brick assembly through the utilization of LEGO-Graph modeling to predict sequential actions. However, existing approaches are class-specific and require significant computational and 3D annotation resources. In this work, we first propose a computationally efficient breadth-first search (BFS) LEGO-Tree structure to model the sequential assembly actions by considering connections between consecutive layers. Based on the LEGO-Tree structure, we then design a class-agnostic tree-transformer framework to predict the sequential assembly actions from the input multi-view images. A major challenge of the sequential brick assembly task is that the step-wise action labels are costly and tedious to obtain in practice. We mitigate this problem by leveraging synthetic-to-real transfer learning. Specifically, our model is first pre-trained on synthetic data with full supervision from the available action labels. We then circumvent the requirement for action labels in the real data by proposing an action-to-silhouette projection that replaces action labels with input image silhouettes for self-supervision. Without any annotation on the real data, our model outperforms existing methods with 3D supervision by 7.8% and 11.3% in mIoU on the MNIST and ModelNet Construction datasets, respectively.
format Preprint
id arxiv_https___arxiv_org_abs_2407_15648
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle TreeSBA: Tree-Transformer for Self-Supervised Sequential Brick Assembly
Guo, Mengqi
Li, Chen
Zhao, Yuyang
Lee, Gim Hee
Computer Vision and Pattern Recognition
Inferring step-wise actions to assemble 3D objects with primitive bricks from images is a challenging task due to complex constraints and the vast number of possible combinations. Recent studies have demonstrated promising results on sequential LEGO brick assembly through the utilization of LEGO-Graph modeling to predict sequential actions. However, existing approaches are class-specific and require significant computational and 3D annotation resources. In this work, we first propose a computationally efficient breadth-first search (BFS) LEGO-Tree structure to model the sequential assembly actions by considering connections between consecutive layers. Based on the LEGO-Tree structure, we then design a class-agnostic tree-transformer framework to predict the sequential assembly actions from the input multi-view images. A major challenge of the sequential brick assembly task is that the step-wise action labels are costly and tedious to obtain in practice. We mitigate this problem by leveraging synthetic-to-real transfer learning. Specifically, our model is first pre-trained on synthetic data with full supervision from the available action labels. We then circumvent the requirement for action labels in the real data by proposing an action-to-silhouette projection that replaces action labels with input image silhouettes for self-supervision. Without any annotation on the real data, our model outperforms existing methods with 3D supervision by 7.8% and 11.3% in mIoU on the MNIST and ModelNet Construction datasets, respectively.
title TreeSBA: Tree-Transformer for Self-Supervised Sequential Brick Assembly
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2407.15648