Neural Contact Fields: Tracking Extrinsic Contact with Tactile Sensing

TL;DR

Neural Contact Fields is a novel method combining neural fields and tactile sensing to track arbitrary external contacts on objects, improving localization accuracy without prior contact assumptions, and supporting complex contact scenarios.

Contribution

It introduces Neural Contact Fields, the first approach capable of tracking multi-modal extrinsic contacts without assuming contact type or shape, and provides a new dataset for this task.

Findings

Successfully localizes multiple contact patches.

Captures contact/no-contact transitions for unseen shapes.

Operates without assumptions on contact geometry.

Abstract

We present Neural Contact Fields, a method that brings together neural fields and tactile sensing to address the problem of tracking extrinsic contact between object and environment. Knowing where the external contact occurs is a first step towards methods that can actively control it in facilitating downstream manipulation tasks. Prior work for localizing environmental contacts typically assume a contact type (e.g. point or line), does not capture contact/no-contact transitions, and only works with basic geometric-shaped objects. Neural Contact Fields are the first method that can track arbitrary multi-modal extrinsic contacts without making any assumptions about the contact type. Our key insight is to estimate the probability of contact for any 3D point in the latent space of object shapes, given vision-based tactile inputs that sense the local motion resulting from the external contact. In experiments, we find that Neural Contact Fields are able to localize multiple contact patches without making any assumptions about the geometry of the contact, and capture contact/no-contact transitions for known categories of objects with unseen shapes in unseen environment configurations. In addition to Neural Contact Fields, we also release our YCB-Extrinsic-Contact dataset of simulated extrinsic contact interactions to enable further research in this area. Project page: https://github.com/carolinahiguera/NCF