# FEWT：用于多模态轮式双臂操作的频率增强小波 Transformer

> 原文链接：https://robotics.hellobike.com/papers/2026-08-31-2509-11109
> 频道：papers | 更新日期：2026-08-31 | 标签：arxiv, cs.RO
> 核心摘要：FEWT缓解视觉语义与高频物理动态表征不匹配，并在仿真双臂插入和真实操作任务中验证。

# FEWT：用于多模态轮式双臂操作的频率增强小波 Transformer

::: v-pre
> 原文链接：[FEWT: Frequency‐Enhanced Wavelet‐Based Transformer for Multimodal Wheeled Bimanual Manipulation](https://arxiv.org/abs/2509.11109)（arXiv 2509.11109 · 发布日期：2026-08-31）

**FEWT: Frequency‐Enhanced Wavelet‐Based Transformer for Multimodal Wheeled Bimanual Manipulation**

FEWT缓解视觉语义与高频物理动态表征不匹配，并在仿真双臂插入和真实操作任务中验证。

## 核心要点

- 研究面向轮式双臂操作中空间视觉语义与局部高频物理动态的表征不匹配问题。
- FEWT结合FE-EMA与TS-DWT，显式提取并对齐多尺度特征，提升时频重校准兼容性。
- 系统将自研STF触觉织物集成到末端执行器，融合接触应力、本体和底盘运动信息。
- 实验显示FEWT在仿真双臂插入任务中较action chunking with transformers基线提高成功率，真实任务中也有性能提升。

**验证范围与局限**：验证范围包括仿真双臂插入任务，以及复杂真实移动和桌面操作任务。

## 英文摘要 (Abstract)

ABSTRACT Embodied intelligence bridges the physical world and information spaces, with robots demonstrating immense potential through imitation learning algorithms. In this study, a custom‐built wheeled bimanual robotic platform equipped with an exoskeleton‐style teleoperation system was utilized to realize intuitive remote manipulation and the efficient collection of anthropomorphic action data. To overcome the representation mismatch between spatial visual semantics and localized high‐frequency physical dynamics, we propose a lightweight frequency‐aligned imitation‐learning framework, termed the frequency‐enhanced wavelet‐based transformer (FEWT). FEWT integrates two primary modules: Frequency‐enhanced efficient multiscale attention (FE‐EMA) and time‐series discrete wavelet transform (TS‐DWT) to explicitly extract and align multiscale features, improving the compatibility between spatial visual representations and temporal‐frequency recalibration. Crucially, for real‐world deployment, this framework is further extended into a multimodal system by seamlessly integrating a self‐developed smart tactile fabric (STF) sensor into the physical end‐effectors, providing local contact‐stress information that complements proprioceptive and chassis‐motion cues in the shared multimodal representation. Experimental evaluations demonstrate that the core FEWT architecture significantly improves the success rate over the widely used action chunking with transformers baseline, particularly during the most challenging phases of simulated bimanual insertion tasks. Furthermore, in complex real‐world mobile and desktop manipulation tasks, the full STF‐enhanced system effectively adapts to microscopic dynamic perturbations, yielding substantial performance enhancements.

作者：Jiaxin Huang, Hanyu Liu, Yunsheng Ma, Jian Shen, Yilin Zheng, Jiayi Wen, Zhigong Song
:::

## 原始参考来源 (Sources)

- [arXiv 2509.11109：FEWT: Frequency‐Enhanced Wavelet‐Based Transformer for Multimodal Wheeled Bimanual Manipulation](https://arxiv.org/abs/2509.11109)