Skip to main content Skip to main navigation

Publikation

Investigating Active Sampling for Hardness Classification with Vision-Based Tactile Sensors

Junyi Chen; Alap Kshirsagar; Frederik Heller; Mario Gomez Andreu; Boris Belousov; Tim Schneider; Lisa Pui Yee Lin; Katja Doerschner; Knut Drewing; Jan Peters
In: Computing Research Repository eprint Journal (CoRR), Vol. abs/2505.13231, Pages 1-7, arXiv, 2025.

Zusammenfassung

One of the most important object properties that humans and robots perceive through touch is hardness. This paper investigates information-theoretic active sampling strate- gies for sample-efficient hardness classification with vision- based tactile sensors. We evaluate three probabilistic classifier models and two model-uncertainty-based sampling strategies on a robotic setup as well as on a previously published dataset of samples collected by human testers. Our findings indicate that the active sampling approaches, driven by uncertainty metrics, surpass a random sampling baseline in terms of accuracy and stability. Additionally, while in our human study, the participants achieve an average accuracy of 48.00%, our best approach achieves an average accuracy of 88.78% on the same set of objects, demonstrating the effectiveness of vision-based tactile sensors for object hardness classification.

Weitere Links