The article explores the use of optical tactile sensors, specifically a see-through-your-skin (STS) variant, in robotic manipulation. This type of sensor combines visual and tactile modes, which can be leveraged for visuotactile sensing. The authors test this approach with imitation learning for tasks that involve contact-rich manipulation, such as opening and closing doors. The results highlight the importance of tactile sensing for imitation learning, both for data collection to allow force matching, and for policy execution to allow accurate task feedback.
Publication date: 3 Nov 2023
Project Page: Not provided
Paper: https://arxiv.org/pdf/2311.01248
