LighTN: Light-Weight Transformer Network for Performance-Overhead Tradeoff in Point Cloud Downsampling

doi:10.1109/TMM.2023.3318073

	LighTN: Light-Weight Transformer Network for Performance-Overhead Tradeoff in Point Cloud Downsampling
	Wang, Xu 1,2; Jin, Yi 1,2; Cen, Yigang 1,2; Wang, Tao 1,2; Tang, Bowen 3; Li, Yidong 1,2
	2025
发表期刊	IEEE TRANSACTIONS ON MULTIMEDIA
ISSN	1520-9210
卷号	27 页码:832-847
摘要	Downsampling is a crucial task for processing large scale and/or dense point clouds with limited resources. Owing to the development of deep learning, approaches of task-oriented point cloud downsampling have significant performance gains in preserving geometric information. However, most downsamling methods are limited by the disordered and unstructured point cloud data, making it difficult to continually improve the performance. To address this issue, we propose a light-weight Transformer network (LighTN) for the task-oriented point cloud downsampling as an end-to-end solution. In LighTN, we design an energy-efficient and permutation invariant single-head self-correlation module to extract refined global geometric features. Moreover, we present a novel sampling loss function to guide LighTN to focus on critical point cloud regions with more uniform distributions and prominent point coverage. Extensive experiments on classification, registration, and reconstruction tasks demonstrate that LighTN can achieve the state-of-the-art performance-overhead tradeoff and high-quality qualitative results.
关键词	Task analysis Point cloud compression Transformers Feature extraction Voltage transformers Three-dimensional displays Deep learning Point cloud deep learning downsampling transformer energy-efficient
DOI	10.1109/TMM.2023.3318073
收录类别	SCI
语种	英语
资助项目	National Natural Science Foundation of China[61972030] ; National Key Research and Development Program of China[2022YFB3103505] ; State Scholarship Fund from the China Scholarship Council
WOS研究方向	Computer Science ; Telecommunications
WOS类目	Computer Science, Information Systems ; Computer Science, Software Engineering ; Telecommunications
WOS记录号	WOS:001428064500037
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
引用统计
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/40703
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Jin, Yi; Cen, Yigang
作者单位	1.Beijing Jiaotong Univ, KeyLaboratory Big Data & Artificial Intelligence T, Minist Educ, Beijing 100044, Peoples R China 2.Beijing Jiaotong Univ, Sch Comp & Informat Technol, Beijing 100044, Peoples R China 3.Chinese Acad Sci, Inst Comp Technol, Comp Architecture, Beijing 100190, Peoples R China
推荐引用方式 GB/T 7714	Wang, Xu,Jin, Yi,Cen, Yigang,et al. LighTN: Light-Weight Transformer Network for Performance-Overhead Tradeoff in Point Cloud Downsampling[J]. IEEE TRANSACTIONS ON MULTIMEDIA,2025,27:832-847.
APA	Wang, Xu,Jin, Yi,Cen, Yigang,Wang, Tao,Tang, Bowen,&Li, Yidong.(2025).LighTN: Light-Weight Transformer Network for Performance-Overhead Tradeoff in Point Cloud Downsampling.IEEE TRANSACTIONS ON MULTIMEDIA,27,832-847.
MLA	Wang, Xu,et al."LighTN: Light-Weight Transformer Network for Performance-Overhead Tradeoff in Point Cloud Downsampling".IEEE TRANSACTIONS ON MULTIMEDIA 27(2025):832-847.