We are pleased to announce that our paper,  “Development of a Sensor Suit for Gait Environment Detection Using Noncontact Sensors and Integrated Model” has been published in the IEEE Transactions on Robotics.

This study proposes EIT-GNN, a graph-structured reconstruction framework that enables super-resolution tactile sensing for electrical impedance tomography on arbitrarily shaped robotic surfaces. By combining a transformer encoder with a graph convolutional network, the method achieves accurate and generalized tactile reconstruction, outperforming existing models and demonstrating effective control on complex large-area sensors such as a face-shaped robotic surface.