RePL: Pseudo-label Refinement for Semi-supervised LiDAR Semantic Segmentation

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Kwon, Donghyeon, Park, Taegyu, Kwak, Suha
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866913015446634496
author Kwon, Donghyeon
Park, Taegyu
Kwak, Suha
author_facet Kwon, Donghyeon
Park, Taegyu
Kwak, Suha
contents Semi-supervised learning for LiDAR semantic segmentation often suffers from error propagation and confirmation bias caused by noisy pseudo-labels. To tackle this chronic issue, we introduce RePL, a novel framework that enhances pseudo-label quality by identifying and correcting potential errors in pseudo-labels through masked reconstruction, along with a dedicated training strategy. We also provide a theoretical analysis demonstrating the condition under which the pseudo-label refinement is beneficial, and empirically confirm that the condition is mild and clearly met by RePL. Extensive evaluations on the nuScenes-lidarseg and SemanticKITTI datasets show that RePL improves pseudo-label quality a lot and, as a result, achieves the state of the art in LiDAR semantic segmentation.
format Preprint
id arxiv_https___arxiv_org_abs_2604_06825
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle RePL: Pseudo-label Refinement for Semi-supervised LiDAR Semantic Segmentation
Kwon, Donghyeon
Park, Taegyu
Kwak, Suha
Computer Vision and Pattern Recognition
Semi-supervised learning for LiDAR semantic segmentation often suffers from error propagation and confirmation bias caused by noisy pseudo-labels. To tackle this chronic issue, we introduce RePL, a novel framework that enhances pseudo-label quality by identifying and correcting potential errors in pseudo-labels through masked reconstruction, along with a dedicated training strategy. We also provide a theoretical analysis demonstrating the condition under which the pseudo-label refinement is beneficial, and empirically confirm that the condition is mild and clearly met by RePL. Extensive evaluations on the nuScenes-lidarseg and SemanticKITTI datasets show that RePL improves pseudo-label quality a lot and, as a result, achieves the state of the art in LiDAR semantic segmentation.
title RePL: Pseudo-label Refinement for Semi-supervised LiDAR Semantic Segmentation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2604.06825