Learning from Mistakes: Self-Regularizing Hierarchical Representations in Point Cloud Semantic Segmentation

TL;DR

This paper introduces LEAK, a self-regularizing hierarchical learning method for LiDAR semantic segmentation that improves accuracy, balance, and convergence by leveraging class error clustering and feature alignment.

Contribution

The paper proposes a novel coarse-to-fine learning framework that enhances segmentation performance without architectural changes by using class error clustering and feature regularization.

Findings

Achieves state-of-the-art results across multiple architectures and datasets.

Ensures more balanced class-wise segmentation results.

Speeds up model convergence during training.

Abstract

Recent advances in autonomous robotic technologies have highlighted the growing need for precise environmental analysis. LiDAR semantic segmentation has gained attention to accomplish fine-grained scene understanding by acting directly on raw content provided by sensors. Recent solutions showed how different learning techniques can be used to improve the performance of the model, without any architectural or dataset change. Following this trend, we present a coarse-to-fine setup that LEArns from classification mistaKes (LEAK) derived from a standard model. First, classes are clustered into macro groups according to mutual prediction errors; then, the learning process is regularized by: (1) aligning class-conditional prototypical feature representation for both fine and coarse classes, (2) weighting instances with a per-class fairness index. Our LEAK approach is very general and can be seamlessly applied on top of any segmentation architecture; indeed, experimental results showed that it enables state-of-the-art performances on different architectures, datasets and tasks, while ensuring more balanced class-wise results and faster convergence.