Design and Development of Tailoring Tools for the Fashion Industry

The integration of advanced technologies in the fashion industry has revolutionized customization, enabling precise garment segmentation and personalized design solutions. This paper presents a novel approach leveraging saliency map extraction and deep learning models for cloth segmentation and style transfer, addressing the limitations of current virtual try-on methods in handling diverse garment types and intricate details. Using ResNet as the core architecture, our model achieved an accuracy of 90.65% , IoU value of 0.9873 and Dice Coefficient of 0.9936, outperforming state-of-the-art networks such as UNet and EfficientNet. By incorporating saliency maps, the proposed system ensures accurate garment extraction, seamless style transfer, and inclusivity across a wide spectrum of clothing options. These advancements bridge the gap between traditional fashion and digital personalization, providing a robust solution for modern applications in virtual try-ons and custom-tailoring.