Class-Level Confidence Based 3D Semi-Supervised Learning

TL;DR

This paper introduces a novel class-level confidence based semi-supervised learning method for 3D data that effectively handles class imbalance, improving performance in classification and detection tasks.

Contribution

The paper proposes a new 3D SSL approach using class-level confidence, dynamic thresholding, and re-sampling strategies to address data imbalance issues.

Findings

Outperforms state-of-the-art methods in 3D SSL classification

Achieves significant improvements in 3D detection tasks

Effectively handles class imbalance in 3D semi-supervised learning

Abstract

Recent state-of-the-art method FlexMatch firstly demonstrated that correctly estimating learning status is crucial for semi-supervised learning (SSL). However, the estimation method proposed by FlexMatch does not take into account imbalanced data, which is the common case for 3D semi-supervised learning. To address this problem, we practically demonstrate that unlabeled data class-level confidence can represent the learning status in the 3D imbalanced dataset. Based on this finding, we present a novel class-level confidence based 3D SSL method. Firstly, a dynamic thresholding strategy is proposed to utilize more unlabeled data, especially for low learning status classes. Then, a re-sampling strategy is designed to avoid biasing toward high learning status classes, which dynamically changes the sampling probability of each class. To show the effectiveness of our method in 3D SSL tasks, we conduct extensive experiments on 3D SSL classification and detection tasks. Our method significantly outperforms state-of-the-art counterparts for both 3D SSL classification and detection tasks in all datasets.