发表机构
Tsinghua University; XSpark AI; The University of Hong Kong; Nanyang Technological University（清华大学; XSpark AI; 香港大学; 南洋理工大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
提出BiView-Touch，通过交叉手补全学习双手触觉表征，在HumanTouch数据集上以5%标签实现7.1%和14.1%的相对平衡准确率提升，并引入BVT-20数据集。
AI 中文摘要
双手交互产生同一物理过程的互补触觉视图，然而现有的触觉表征学习大多独立地对两只手进行建模，或仅将其结合用于下游预测，而忽略了双手之间的交叉关系。为了利用这一被忽视的结构，我们提出了BiView-Touch，一个仅基于触觉的框架，该框架从剩余的可见目标手区域和同步的完整对侧手区域补全被掩码的目标手潜在表示。学生编码器配备了几何条件定向解码器，用于预测全视图EMA潜在目标，而时间与布局反事实则鼓励对同步且解剖学组织的源信息保持敏感。受控消融实验和源上下文干预表明，BiView-Touch学习了对时间对齐且解剖学组织的对侧触觉上下文的结构化交叉手依赖，而非仅仅受益于双边输入。在公开的HumanTouch数据集上，其冻结表征在低标签设置下始终优于代表性的自监督基线。仅使用5%的下游标签，BiView-Touch在双侧手腕运动识别上实现了7.1%的相对平衡准确率提升，在基于力的交互阶段识别上实现了14.1%的提升。我们进一步引入了BVT-20，一个包含20个任务的双手触觉数据集，并展示了跨记录会话和预训练语料库的迁移能力，包括向一个留出的双手任务的迁移。我们的代码和数据集细节可在匿名项目页面获取：此https URL。
英文摘要
Bimanual interaction produces complementary tactile views of the same physical process, yet existing tactile representation learning largely models the two hands independently or combines them only for downstream prediction, leaving their cross-hand relationship unexplored. To exploit this overlooked structure, we introduce BiView-Touch, a tactile-only framework that completes masked target-hand latents from the remaining visible target-hand regions and the synchronized full contralateral hand. A student encoder with a geometry-conditioned directional decoder predicts full-view EMA latent targets, while temporal and layout counterfactuals encourage sensitivity to synchronized and anatomically organized source information. Controlled ablations and source-context interventions show that BiView-Touch learns structured cross-hand dependence on temporally aligned and anatomically organized contralateral tactile context, rather than benefiting from bilateral input alone. On the public HumanTouch dataset, its frozen representations consistently outperform representative self-supervised baselines across low-label settings. With only 5\% downstream labels, BiView-Touch achieves relative balanced-accuracy gains of 7.1\% on bilateral wrist-motion recognition and 14.1\% on force-derived interaction-phase recognition. We further introduce BVT-20, a 20-task bilateral tactile dataset, and demonstrate transfer across recording sessions and pretraining corpora, including transfer to a held-out bimanual task. Our code and dataset details are available on the anonymous project page: https://anonymous.4open.science/w/biview-touch-review-site-050C/.
CommentsSubmitted to ICRA 2027; 9 pages, 7 figures
