针对3D数据长尾分布，提出兼顾准确率与均衡性的数据剪枝方法。
Exploring 3D Dataset Pruning
- 基于加权子集近似全量数据风险，解决代表性不足和先验偏差问题。
- 在多个3D数据集上同时提升整体准确率和平均准确率，效果稳定。
- 支持灵活调整指标侧重，适合不同下游任务需求。
数据剪枝在2D图像领域已被广泛研究，用于消除冗余并加速训练，但针对3D数据的专门剪枝方法仍较少。本文研究3D数据剪枝，发现其常见的长尾类分布特性导致在传统评估指标整体准确率（OA）和平均准确率（mAcc）下优化目标天然冲突，使剪枝尤为困难。为此，我们将剪枝建模为用加权子集逼近全数据期望风险，揭示两大关键误差：覆盖误差（代表性不足）与先验不匹配偏差（子集诱导的类别权重与目标指标不一致）。提出感知表征的子集选择方法，结合每类保留配额以保障长尾覆盖，并采用校准软标签与嵌入几何蒸馏实现先验不变的教师监督。保留配额还可作为开关调节OA与mAcc之间的权衡。在多个3D数据集上的大量实验表明，该方法可在多种设置下同时提升两个指标，且能适应不同下游偏好。代码已公开于 https://github.com/XiaohanZhao123/3D-Dataset-Pruning。
原文摘要 · Abstract (English)
Dataset pruning has been widely studied for 2D images to remove redundancy and accelerate training, while particular pruning methods for 3D data remain largely unexplored. In this work, we study dataset pruning for 3D data, where its observed common long-tail class distribution nature make optimization under conventional evaluation metrics Overall Accuracy (OA) and Mean Accuracy (mAcc) inherently conflicting, and further make pruning particularly challenging. To address this, we formulate pruning as approximating the full-data expected risk with a weighted subset, which reveals two key errors: coverage error from insufficient representativeness and prior-mismatch bias from inconsistency between subset-induced class weights and target metrics. We propose representation-aware subset selection with per-class retention quotas for long-tail coverage, and prior-invariant teacher supervision using calibrated soft labels and embedding-geometry distillation. The retention quota also serves as a switch to control the OA-mAcc trade-off. Extensive experiments on 3D datasets show that our method can improve both metrics across multiple settings while adapting to different downstream preferences. Our code is available at https://github.com/XiaohanZhao123/3D-Dataset-Pruning.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。