Seeing Through Touch: Tactile-Driven Visual Localization of Material Regions

TL;DR

This paper introduces a tactile-driven visual localization method that learns local cross-modal features for material segmentation, overcoming dataset limitations and outperforming prior approaches.

Contribution

The work presents a novel dense cross-modal alignment model, new datasets, and a pairing strategy to improve tactile localization of material regions in images.

Findings

Our method significantly outperforms prior visuo-tactile approaches.

New datasets enable robust evaluation of tactile localization.

Material-diversity pairing enhances contextual localization and robustness.

Abstract

We address the problem of tactile localization, where the goal is to identify image regions that share the same material properties as a tactile input. Existing visuo-tactile methods rely on global alignment and thus fail to capture the fine-grained local correspondences required for this task. The challenge is amplified by existing datasets, which predominantly contain close-up, low-diversity images. We propose a model that learns local visuo-tactile alignment via dense cross-modal feature interactions, producing tactile saliency maps for touch-conditioned material segmentation. To overcome dataset constraints, we introduce: (i) in-the-wild multi-material scene images that expand visual diversity, and (ii) a material-diversity pairing strategy that aligns each tactile sample with visually varied yet tactilely consistent images, improving contextual localization and robustness to weak signals. We also construct two new tactile-grounded material segmentation datasets for quantitative evaluation. Experiments on both new and existing benchmarks show that our approach substantially outperforms prior visuo-tactile methods in tactile localization.