Transfer learning for vision-based tactile sensing
This work addresses the challenge of efficient tactile sensing for robotics applications, but it is incremental as it builds on existing vision-based methods with a focus on transfer across sensors.
The paper tackles the problem of reconstructing normal force distribution on a soft optical tactile sensor using images from an internal camera, achieving satisfactory performance with reduced training times and data needs through a calibration-based transfer learning approach.
Due to the complexity of modeling the elastic properties of materials, the use of machine learning algorithms is continuously increasing for tactile sensing applications. Recent advances in deep neural networks applied to computer vision make vision-based tactile sensors very appealing for their high-resolution and low cost. A soft optical tactile sensor that is scalable to large surfaces with arbitrary shape is discussed in this paper. A supervised learning algorithm trains a model that is able to reconstruct the normal force distribution on the sensor's surface, purely from the images recorded by an internal camera. In order to reduce the training times and the need for large datasets, a calibration procedure is proposed to transfer the acquired knowledge across multiple sensors while maintaining satisfactory performance.