# FWL-ToPM实现全波形LiDAR实时感知：8倍加速且F1提升9点

> 原文链接：https://robotics.hellobike.com/papers/2026-11-05-2611-07613-a09f9dfc
> 频道：papers | 更新日期：2026-11-05 | 标签：auto, arxiv, hf-daily-papers
> 核心摘要：针对全波形LiDAR计算冗余与数据稀缺痛点，提出FWL-ToPM剪枝机制与FWLAug增强框架。在Ghost-FWL基准上实现8倍以上实时加速，F1分数较SOTA提升近9点，并开源代码与OOD数据集，推动FWL感知工程化落地。

# FWL-ToPM实现全波形LiDAR实时感知：8倍加速且F1提升9点

::: v-pre
> 本文由 AI 根据公开内容整理，可能有误，请以原文为准：[Towards Real-Time Full-Waveform LiDAR Transformers via Intensity-Guided Token Reduction and Physics-Aware Augmentation](https://arxiv.org/abs/2611.07613)（HF Daily Papers · arXiv 2611.07613）

## 核心内容

全波形LiDAR（FWL）虽能通过完整回波直方图解决雾雨及低反射率物体探测难题，但Transformer处理面临背景Token冗余与标注数据匮乏双重瓶颈。研究团队提出FWL-ToPM与FWLAug两套方案，在Ghost-FWL基准测试中实现超8倍推理加速，同时F1分数较现有SOTA提升近9点，成为首个兼顾实时性与精度提升的FWL Transformer模型，相关代码及分布外（OOD）数据集已开源。

FWL-ToPM利用强度分布选择性剪枝合并Token，剔除低强度背景冗余计算并保留关键波形特征；FWLAug作为正则化增强框架，严格保留单射线飞行时间波形与视锥几何结构以提升泛化能力。实验表明，该组合不仅在标准基准表现优异，在未见场景的OOD测试集中同样验证了鲁棒性，有效解决了FWL数据物理结构利用率低导致的部署难题。

## 关键看点

- **Token剪枝机制**：FWL-ToPM基于强度分布选择性合并Token，剔除低强度背景冗余，实现超8倍推理加速并保持实时吞吐
- **物理感知增强**：FWLAug保留单射线ToF波形与视锥几何结构作为正则化手段，显著提升模型在未见场景下的泛化性能
- **SOTA性能突破**：在Ghost-FWL基准上F1分数较前代SOTA提升近9点，是首个同时实现精度与实时性双增的FWL Transformer
- **开源资产释放**：项目公开完整代码及分布外（OOD）测试数据集，为全波形LiDAR感知算法研究提供标准化验证基础

## 工程与产业观察

该工作证明了利用传感器物理先验（强度分布、波形几何）优化Transformer架构的有效性，为具身智能在恶劣环境下的感知提供了低成本实时解法。开源的OOD数据集填补了FWL泛化评测空白，对机器人3D感知算法从实验室走向非结构化场景部署具有重要参考价值。

## 英文摘要 (Abstract)

LiDAR is a critical sensor for autonomous driving and robotics, yet it remains vulnerable to adverse conditions such as fog, rain, transparent objects, and low-reflectance surfaces. Full-waveform LiDAR (FWL) addresses these limitations by capturing the complete return waveform as a histogram, preserving rich reflection characteristics including multi-path echoes and weak signals. While recent Transformer-based approaches have demonstrated promise on FWL data, two fundamental bottlenecks remain: high computational cost from redundant tokens in low-intensity background regions, and limited generalization due to scarce annotated training data. We address both bottlenecks by exploiting the physical structure inherent to FWL data. First, we propose FWL-ToPM, a token pruning and merging mechanism that leverages intensity distributions to selectively reduce tokens while retaining critical waveform content. Second, we introduce FWLAug, a data augmentation framework designed as a regularization method that preserves waveform structure, retaining the Time-of-Flight waveform of each ray and the frustum geometry, to improve generalization to unseen scenes. Evaluated on the Ghost-FWL benchmark, FWL-ToPM achieves over 8 times speedup at real-time throughput alongside a nearly 9-point F1 gain over the prior state of the art, establishing the first real-time FWL Transformer that simultaneously improves accuracy. FWLAug further improves performance on both the standard benchmark and an out-of-distribution test set of unseen scenes. Together, these methods push FWL-based perception toward practical deployment. The code and the OOD dataset are publicly available and can be accessed via the project page: https://keio-csg.github.io/FWL-ToPM

Hugging Face Daily Papers 热度：1 upvotes
作者：Kotaro Oishi, Kazuma Ikeda, Ryosei Hara, Ryo Yoshida, Mariko Isogawa, Kentaro Yoshioka
:::

## 原始参考来源 (Sources)

- [HF Daily Papers · arXiv 2611.07613：Towards Real-Time Full-Waveform LiDAR Transformers via Intensity-Guided Token Reduction and Physics-Aware Augmentation](https://arxiv.org/abs/2611.07613)