IAN: Instance-Augmented Net for 3D Instance Segmentation

doi:10.1109/LRA.2023.3281905

CORC > 自动化研究所 > 中国科学院自动化研究所 > 智能制造技术与系统研究中心 > 先进制造与自动化

	IAN: Instance-Augmented Net for 3D Instance Segmentation
	Wan, Zihao 1,2; Hu, Jianhua 1; Zhang, Haojian 1; Wang, Yunkuan 1
刊名	IEEE ROBOTICS AND AUTOMATION LETTERS
	2023-07-01
卷号	8 期号:7 页码:4354-4361
关键词	Three-dimensional displays Feature extraction Point cloud compression Solid modeling Semantics Noise measurement Aggregates Deep learning for visual perception RGB-D perception data sets for robotic vision
ISSN号	2377-3766
DOI	10.1109/LRA.2023.3281905
通讯作者	Wang, Yunkuan(yunkuan.wang@ia.ac.cn)
英文摘要	When aggregating local information from neighbors, prevailing 3D instance segmentation backbones only leverage 3D coordinates to find neighboring points without identifying whether these points are from the same object as the query point, which causes the model to gather excessive noisy features. Besides, traditional backbones fail to fully utilize multi-resolution information. Therefore, previous methods have difficulty in segmenting targets in cluttered scenes. To tackle these issues, we propose Instance-Augmented Net (IAN). The keys to our approach are Instance-Augmented Block (IAB), Instance-Augmented Upsampler (IAU), and Attentive Fusion (AF). In IAB, for each foreground point, we leverage its instance information to filter out noisy neighbors from other objects. We also propose IAU to apply this instance-augmented strategy to the upsampling process. Furthermore, to retain comprehensive information, we upsample multi-resolution feature maps and adopt attention generated by AF to fuse them. Notably, by encoding neighborhood information, AF can generate attention at point-level adaptively. Moreover, to further test the generality of models, we present Clutter and Occlusion (CAO), a new 3D instance segmentation dataset tailored for robotic grasping tasks. Extensive experiments on S3DIS, ScanNet and CAO show the effectiveness of our IAN.
WOS研究方向	Robotics
语种	英语
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
WOS记录号	WOS:001012840800007
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/53534]
专题	智能制造技术与系统研究中心_先进制造与自动化
通讯作者	Wang, Yunkuan
作者单位	1.Chinese Acad Sci, Inst Automat, Beijing 100190, Peoples R China 2.Univ Chinese Acad Sci, Sch Artificial Intelligence, Beijing 100049, Peoples R China
推荐引用方式 GB/T 7714	Wan, Zihao,Hu, Jianhua,Zhang, Haojian,et al. IAN: Instance-Augmented Net for 3D Instance Segmentation[J]. IEEE ROBOTICS AND AUTOMATION LETTERS,2023,8(7):4354-4361.
APA	Wan, Zihao,Hu, Jianhua,Zhang, Haojian,&Wang, Yunkuan.(2023).IAN: Instance-Augmented Net for 3D Instance Segmentation.IEEE ROBOTICS AND AUTOMATION LETTERS,8(7),4354-4361.
MLA	Wan, Zihao,et al."IAN: Instance-Augmented Net for 3D Instance Segmentation".IEEE ROBOTICS AND AUTOMATION LETTERS 8.7(2023):4354-4361.