A Method of Enhancing Silk Yarn‐Dyed Interwoven Fabric Color Prediction Through CNN Approaches

Color is a fundamental characteristic in textiles, particularly in interwoven mixed‐color fabrics, where color appearance manifests as a nonlinear and complex phenomenon influenced by multiple factors, including yarn material, yarn color, weave structure, and the weaving process. Traditional mathematical models typically exhibit limitations in accurately capturing and managing this complexity. This study presents a deep learning approach utilizing convolutional neural networks (CNNs) to quantitatively analyze and predict the color characteristics of silk interwoven mixed‐color fabrics. The primary objective of this research was to enhance the accuracy and stability of color predictions while enabling visual representations of the predicted outcomes. A comprehensive dataset of 1350 samples was established to train and validate the CNN model in color prediction tasks. Evaluation of the model's performance yielded a high predictive accuracy, with an average color difference (∆ E 00 ) of 0.96 between predicted and actual colors—a value close to the threshold of human visual indiscernibility. The results demonstrate that the CNN model offers precise and generalized color prediction capabilities, providing an intuitive visualization of predicted colors, with significant implications for textile design and production processes.