PointLoRA: Low-Rank Adaptation with Token Selection for Point Cloud Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wang, Song, Liu, Xiaolu, Kong, Lingdong, Xu, Jianyun, Hu, Chunyong, Fang, Gongfan, Li, Wentong, Zhu, Jianke, Wang, Xinchao
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913860770856960
author Wang, Song
Liu, Xiaolu
Kong, Lingdong
Xu, Jianyun
Hu, Chunyong
Fang, Gongfan
Li, Wentong
Zhu, Jianke
Wang, Xinchao
author_facet Wang, Song
Liu, Xiaolu
Kong, Lingdong
Xu, Jianyun
Hu, Chunyong
Fang, Gongfan
Li, Wentong
Zhu, Jianke
Wang, Xinchao
contents Self-supervised representation learning for point cloud has demonstrated effectiveness in improving pre-trained model performance across diverse tasks. However, as pre-trained models grow in complexity, fully fine-tuning them for downstream applications demands substantial computational and storage resources. Parameter-efficient fine-tuning (PEFT) methods offer a promising solution to mitigate these resource requirements, yet most current approaches rely on complex adapter and prompt mechanisms that increase tunable parameters. In this paper, we propose PointLoRA, a simple yet effective method that combines low-rank adaptation (LoRA) with multi-scale token selection to efficiently fine-tune point cloud models. Our approach embeds LoRA layers within the most parameter-intensive components of point cloud transformers, reducing the need for tunable parameters while enhancing global feature capture. Additionally, multi-scale token selection extracts critical local information to serve as prompts for downstream fine-tuning, effectively complementing the global context captured by LoRA. The experimental results across various pre-trained models and three challenging public datasets demonstrate that our approach achieves competitive performance with only 3.43% of the trainable parameters, making it highly effective for resource-constrained applications. Source code is available at: https://github.com/songw-zju/PointLoRA.
format Preprint
id arxiv_https___arxiv_org_abs_2504_16023
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle PointLoRA: Low-Rank Adaptation with Token Selection for Point Cloud Learning
Wang, Song
Liu, Xiaolu
Kong, Lingdong
Xu, Jianyun
Hu, Chunyong
Fang, Gongfan
Li, Wentong
Zhu, Jianke
Wang, Xinchao
Computer Vision and Pattern Recognition
Self-supervised representation learning for point cloud has demonstrated effectiveness in improving pre-trained model performance across diverse tasks. However, as pre-trained models grow in complexity, fully fine-tuning them for downstream applications demands substantial computational and storage resources. Parameter-efficient fine-tuning (PEFT) methods offer a promising solution to mitigate these resource requirements, yet most current approaches rely on complex adapter and prompt mechanisms that increase tunable parameters. In this paper, we propose PointLoRA, a simple yet effective method that combines low-rank adaptation (LoRA) with multi-scale token selection to efficiently fine-tune point cloud models. Our approach embeds LoRA layers within the most parameter-intensive components of point cloud transformers, reducing the need for tunable parameters while enhancing global feature capture. Additionally, multi-scale token selection extracts critical local information to serve as prompts for downstream fine-tuning, effectively complementing the global context captured by LoRA. The experimental results across various pre-trained models and three challenging public datasets demonstrate that our approach achieves competitive performance with only 3.43% of the trainable parameters, making it highly effective for resource-constrained applications. Source code is available at: https://github.com/songw-zju/PointLoRA.
title PointLoRA: Low-Rank Adaptation with Token Selection for Point Cloud Learning
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2504.16023