Deep Learning-Enabled colour Matching in Textile Design: Effects on colour Matching Accuracy, Visual Communication Effectiveness, and Design Performance
Description
Deep learning-based colour matching in textiles brings about revolutionary changes in textile design by enhancing accuracy, communication, efficiency, and decision-making regarding colours. The study explores deep learning-enabled colour matching in textile design and its effects on colour-matching accuracy, visual communication effectiveness, and design performance. A mixed-methods research design using purposive sampling was utilized. Quantitative data were collected from 281 participants, including undergraduate and postgraduate students and professional textile/fashion designers, through a structured online questionnaire. Qualitative data were collected through interviews with 43 participants selected from the quantitative sample through online interviews. Data were analyzed using SPSS, SmartPLS, and NVivo. Findings indicate that Deep Learning Enabled Colour Matching (DLE) significantly improved Colour Matching Accuracy (CMA) (β = 0.236, p < 0.001) and Visual Communication Effectiveness (VCE) (β = 0.218, p < 0.001), while CMA significantly influenced VCE (β = 0.270, p < 0.001). CMA significantly mediated the DLE-VCE relationship (β = 0.063, p = 0.005). VCE significantly improved textile design performance (β = 0.422, p < 0.001). Design expertise significantly moderated the DLE-CMA relationship (β = 0.271, p < 0.001). Qualitative findings highlight improved accuracy, communication, productivity, and the continuing importance of design expertise. This study examines deep learning-enabled colour matching, with accuracy as a mediator and design expertise as a moderator linking communication and textile design processes. The use of DLE colour matching by textile designers requires integration with human judgment, proper testing of the systems used, and training for accuracy and effectiveness.