Physics-Driven Learning Framework for Tomographic Tactile Sensing

Abstract

Electrical impedance tomography (EIT) provides an attractive solution for large-area tactile sensing due to its minimal wiring and geometric flexibility, but its nonlinear inverse problem often leads to severe artifacts and inaccurate reconstruction. This work presents PhyDNN, a physics-driven deep reconstruction framework that embeds the EIT forward model directly into the learning objective. By jointly minimizing the discrepancy between predicted and ground-truth conductivity maps and enforcing consistency with the forward PDE, PhyDNN reduces the opaque nature of deep networks and improves both physical plausibility and generalization. To enable efficient backpropagation, we design a differentiable forward-operator network that accurately approximates the nonlinear EIT response, allowing fast physics-guided training. Extensive simulations and real tactile experiments on a 16-electrode soft sensor demonstrate that PhyDNN consistently outperforms NOSER, TV, and standard DNNs in reconstructing contact shape and location. The proposed method yields fewer artifacts, sharper boundaries, and higher quantitative scores, demonstrating its effectiveness for high-quality tomographic tactile sensing.