Generating training data for estimating material property parameter of fabric and estimating material property parameter of fabric

A method of generating training data of an artificial neural network, the method comprising:

• receiving a first number of material property parameters of first different sample fabrics;
• generating a second number of material property parameters of second different sample fabrics by applying a generation model to upsample the first number of the material property parameters, the second number being greater than the first number;
• performing simulation to generate three-dimensional (3D) shapes of the second different sample fabrics corresponding to the second number of material property parameters, the shapes of the fabrics including portions of the fabrics hanging from a 3D geometric object;
• and
• generating training data including the generated 3D shapes of the simulated fabrics and the second number of material property parameters.

A non-transitory computer-readable storage medium storing instructions thereon, the instructions when executed by a processor cause the processor to:

• receive a three-dimensional (3D) shape of fabric with at least a portion of the fabric placed on a 3D geometric object and another portion of the fabric hanging from the 3D geometric object;
• process the 3D shape of the fabric by a trained artificial neural network to estimate a material property parameter of the fabric, the artificial neural network trained using training data that includes a second number of material property parameters of second sample fabrics that are generated by applying a generation model to upsample a first number of material property parameters of first sample fabrics, the second number being greater than the first number;
• and
• provide the estimated material property parameter of the fabric.

A non-transitory computer-readable storage medium storing instructions thereon, the instructions representing an artificial neural network trained by:

• receiving a first number of material property parameters of first different sample fabrics;
• generating a second number of material property parameters of second different sample fabrics by applying a generation model to upsample the first number of the material property parameters, the second number being greater than the first number;
• performing simulation to generate three-dimensional (3D) shapes of the second different sample fabrics corresponding to the second number of material property parameters, the shapes of the fabrics including portions of the fabrics hanging from a 3D geometric object;
• including the generated 3D shapes of the simulated fabrics and the second number of material property parameters in training data;
• and
• training the artificial neural network using the training data.