APoX-M: Accelerate deep point cloud analysis via adaptive graph construction

doi:10.1016/j.vlsi.2024.102313

	APoX-M: Accelerate deep point cloud analysis via adaptive graph construction
	Dai, Lei 1,2; Liang, Shengwen 1,2,3; Wang, Ying 2,3,4,5; Li, Huawei 1,2; Li, Xiaowei 1,2,3
	2025-03-01
发表期刊	INTEGRATION-THE VLSI JOURNAL
ISSN	0167-9260
卷号	101 页码:10
摘要	Graph-based deep learning point cloud processing has gained increasing popularity but its performance is dragged by the dominating graph construction (GC) phase with irregular computation and memory access. Existing works that accelerate GC by tailoring architecture fora single GC algorithm fail to maintain efficiency because they neglect the best GC algorithm variation incurred by the point-cloud density variation in changing scenarios. Therefore, we propose APoX-M, a unified architecture with an adaptive GC scheme that can identify the optimum GC approach according to the point cloud variation. We also provide better memory management and scheduling optimizations for better performance. Experiments indicate that APoX-M achieves higher performance and energy efficiency over existing accelerators.
关键词	K nearest neighbor Approximate nearest neighbor Point cloud Graph construction
DOI	10.1016/j.vlsi.2024.102313
收录类别	SCI
语种	英语
资助项目	National Key Research and Development Program of China[2018AAA0102705] ; National Natural Science Foundation of China (NSFC) , China[62202453] ; National Natural Science Foundation of China (NSFC) , China[62090024] ; National Natural Science Foundation of China (NSFC) , China[61876173] ; China Postdoctoral Science Foundation[2022M713207]
WOS研究方向	Computer Science ; Engineering
WOS类目	Computer Science, Hardware & Architecture ; Engineering, Electrical & Electronic
WOS记录号	WOS:001382288400001
出版者	ELSEVIER
引用统计
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/41071
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Liang, Shengwen
作者单位	1.Chinese Acad Sci, Inst Comp Technol, SKLP, 6 Zhongguancun South St, Beijing 100190, Peoples R China 2.Univ Chinese Acad Sci, 80 Zhongguancun East Rd, Beijing 100190, Peoples R China 3.Zhongguancun Natl Lab, Beijing 100190, Peoples R China 4.Chinese Acad Sci, Inst Comp Technol, CICS, 6 Zhongguancun South St, Beijing 100190, Peoples R China 5.Zhejiang Lab, Hangzhou, Zhejiang, Peoples R China
推荐引用方式 GB/T 7714	Dai, Lei,Liang, Shengwen,Wang, Ying,et al. APoX-M: Accelerate deep point cloud analysis via adaptive graph construction[J]. INTEGRATION-THE VLSI JOURNAL,2025,101:10.
APA	Dai, Lei,Liang, Shengwen,Wang, Ying,Li, Huawei,&Li, Xiaowei.(2025).APoX-M: Accelerate deep point cloud analysis via adaptive graph construction.INTEGRATION-THE VLSI JOURNAL,101,10.
MLA	Dai, Lei,et al."APoX-M: Accelerate deep point cloud analysis via adaptive graph construction".INTEGRATION-THE VLSI JOURNAL 101(2025):10.