Learning Physics Properties of Fabrics and Garments with a Physics Similarity Neural Network
This work addresses the challenge of accurately modeling fabric and garment physics for applications in computer graphics or robotics, representing a domain-specific incremental advance.
The paper tackles the problem of predicting physics parameters like bending stiffness for real fabrics and garments by learning physics similarities from simulations, achieving improvements of 34% for fabrics and 68% for garments over state-of-the-art methods.
In this paper, we propose to predict the physics parameters of real fabrics and garments by learning their physics similarities between simulated fabrics via a Physics Similarity Network (PhySNet). For this, we estimate wind speeds generated by an electric fan and the area weight to predict bending stiffness of simulated and real fabrics and garments. We found that PhySNet coupled with a Bayesian optimiser can predict physics parameters and improve the state-of-art by 34%for real fabrics and 68% for real garments.