Consistency-aware Self-Training for Iterative-based Stereo Matching

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhou, Jingyi, Ye, Peng, Zhang, Haoyu, Yuan, Jiakang, Qiang, Rao, YangChenXu, Liu, Cailin, Wu, Xu, Feng, Chen, Tao
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917972415610880
author Zhou, Jingyi
Ye, Peng
Zhang, Haoyu
Yuan, Jiakang
Qiang, Rao
YangChenXu, Liu
Cailin, Wu
Xu, Feng
Chen, Tao
author_facet Zhou, Jingyi
Ye, Peng
Zhang, Haoyu
Yuan, Jiakang
Qiang, Rao
YangChenXu, Liu
Cailin, Wu
Xu, Feng
Chen, Tao
contents Iterative-based methods have become mainstream in stereo matching due to their high performance. However, these methods heavily rely on labeled data and face challenges with unlabeled real-world data. To this end, we propose a consistency-aware self-training framework for iterative-based stereo matching for the first time, leveraging real-world unlabeled data in a teacher-student manner. We first observe that regions with larger errors tend to exhibit more pronounced oscillation characteristics during model prediction.Based on this, we introduce a novel consistency-aware soft filtering module to evaluate the reliability of teacher-predicted pseudo-labels, which consists of a multi-resolution prediction consistency filter and an iterative prediction consistency filter to assess the prediction fluctuations of multiple resolutions and iterative optimization respectively. Further, we introduce a consistency-aware soft-weighted loss to adjust the weight of pseudo-labels accordingly, relieving the error accumulation and performance degradation problem due to incorrect pseudo-labels. Extensive experiments demonstrate that our method can improve the performance of various iterative-based stereo matching approaches in various scenarios. In particular, our method can achieve further enhancements over the current SOTA methods on several benchmark datasets.
format Preprint
id arxiv_https___arxiv_org_abs_2503_23747
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Consistency-aware Self-Training for Iterative-based Stereo Matching
Zhou, Jingyi
Ye, Peng
Zhang, Haoyu
Yuan, Jiakang
Qiang, Rao
YangChenXu, Liu
Cailin, Wu
Xu, Feng
Chen, Tao
Computer Vision and Pattern Recognition
Iterative-based methods have become mainstream in stereo matching due to their high performance. However, these methods heavily rely on labeled data and face challenges with unlabeled real-world data. To this end, we propose a consistency-aware self-training framework for iterative-based stereo matching for the first time, leveraging real-world unlabeled data in a teacher-student manner. We first observe that regions with larger errors tend to exhibit more pronounced oscillation characteristics during model prediction.Based on this, we introduce a novel consistency-aware soft filtering module to evaluate the reliability of teacher-predicted pseudo-labels, which consists of a multi-resolution prediction consistency filter and an iterative prediction consistency filter to assess the prediction fluctuations of multiple resolutions and iterative optimization respectively. Further, we introduce a consistency-aware soft-weighted loss to adjust the weight of pseudo-labels accordingly, relieving the error accumulation and performance degradation problem due to incorrect pseudo-labels. Extensive experiments demonstrate that our method can improve the performance of various iterative-based stereo matching approaches in various scenarios. In particular, our method can achieve further enhancements over the current SOTA methods on several benchmark datasets.
title Consistency-aware Self-Training for Iterative-based Stereo Matching
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2503.23747