Generating Inverse Feature Space for Class Imbalance in Point Cloud Semantic Segmentation

分割 计算机科学 人工智能 点云 班级(哲学) 特征(语言学) 模式识别(心理学) 点(几何) 空格(标点符号) 特征向量 数学 几何学 语言学 操作系统 哲学
作者
Jiawei Han,Kaiqi Liu,Wei Li,Feng Zhang,Xiang‐Gen Xia
出处
期刊:IEEE Transactions on Pattern Analysis and Machine Intelligence [IEEE Computer Society]
卷期号:47 (7): 5778-5793 被引量:30
标识
DOI:10.1109/tpami.2025.3553051
摘要

Point cloud semantic segmentation can enhance the understanding of the production environment and is a crucial component of vision tasks. The efficacy and generalization prowess of deep learning-based segmentation models are inherently contingent upon the quality and nature of the data employed in their training. However, it is often challenging to obtain data with inter-class balance, and training an intelligent segmentation network with the imbalanced data may cause cognitive bias. In this paper, a network framework InvSpaceNet is proposed, which generates an inverse feature space to alleviate the cognitive bias caused by imbalanced data. Specifically, we design a dual-branch training architecture that combines the superior feature representations derived from instance-balanced sampling data with the cognitive corrections introduced by the proposed inverse sampling data. In the inverse feature space of the point cloud generated by the auxiliary branch, the central points aggregated by class are constrained by the contrastive loss. To refine the class cognition in the inverse feature space, features are used to generate point cloud class prototypes through momentum update. These class prototypes from the inverse space are utilized to generate feature maps and structure maps that are aligned with the positive feature space of the main branch segmentation network. The training of the main branch is dynamically guided through gradients back propagated from different losses. Extensive experiments conducted on four large benchmarks (i.e., S3DIS, ScanNet v2, Toronto-3D, and SemanticKITTI) demonstrate that the proposed method can effectively mitigate point cloud imbalance issues and improve segmentation performance.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
个性冰萍完成签到 ,获得积分10
刚刚
刚刚
1秒前
2秒前
CipherSage的应助被yep采纳,获得10
2秒前
夏天就是桃子味完成签到,获得积分10
2秒前
风吹麦浪的应助被香蕉寒珊采纳,获得10
2秒前
香蕉觅云的应助被允蓥采纳,获得10
2秒前
可爱的函函的应助被Moonber采纳,获得10
3秒前
4秒前
5秒前
小东西发布了新的文献求助10
6秒前
6秒前
aaaaa发布了新的文献求助10
6秒前
6秒前
秋风的应助被dm采纳,获得20
6秒前
7秒前
CC发布了新的文献求助30
7秒前
负责以山完成签到 ,获得积分10
9秒前
Shuo完成签到,获得积分10
9秒前
freeze完成签到,获得积分10
9秒前
9秒前
打打的应助被俭朴采纳,获得10
10秒前
dukui完成签到,获得积分10
10秒前
CipherSage的应助被怡然的小虾米采纳,获得10
10秒前
hhh发布了新的文献求助10
10秒前
科研通AI65的应助被yao采纳,获得10
10秒前
焦粉蛋挞发布了新的文献求助10
10秒前
顾矜的应助被YANGYINGSHUO采纳,获得10
10秒前
布豪办完成签到,获得积分10
11秒前
11秒前
ililiiris发布了新的文献求助10
11秒前
Lisa的应助被辛勤电源采纳,获得10
12秒前
12秒前
wang完成签到 ,获得积分10
14秒前
LPhy_Z完成签到,获得积分10
14秒前
七月流火的应助被SamYang采纳,获得10
15秒前
允蓥发布了新的文献求助10
15秒前
bbb完成签到,获得积分10
15秒前
16秒前
高分求助中
(应助此贴封号)【重要!!请各用户(尤其是新用户)详细阅读】【科研通的精品贴汇总】 10000
拉伸高分子熔体中缠结行为的模拟和小角中子散射研究 1700
The World Atlas of Wine 9th edition 1000
奈特基础组织学:配套组织病理学(第4版)(Netter's Essential Histology With Correlated Histopathology) 800
八旗制度の研究 600
Improving the wear resistance of vehicle gear oil by oil soluble LaF3 nanoparticles and typical additives of gear oil 500
Boundaries and Bodhisattvas 500
热门求助领域 (近24小时)
化学 材料科学 医学 生物 纳米技术 计算机科学 工程类 有机化学 化学工程 生物化学 物理 复合材料 内科学 催化作用 光电子学 人工智能 细胞生物学 无机化学 物理化学 数学
热门帖子
关注 科研通微信公众号,转发送积分 8183204
求助须知:如何正确求助?哪些是违规求助? 9572482
关于积分的说明 21825940
捐赠科研通 7780074
什么是DOI,文献DOI怎么找? 3433945
关于科研通互助平台的介绍 2558060
邀请新用户注册赠送积分活动 2480166