The application of hierarchical perception technology based on deep learning in 3D fashion design

Abstract

In recent years, 3D fashion design has been relied on for improving attire fashion, design, and presentation with fewer flaws and better visualization. This aids consumers in providing visualized recommendations on modifications, suggestions, and customized attire designs. Considering the influence of automation and intelligent processing in the fashion designing industry, it introduces a Flaw Detection Method in 3D Representation (FDM-3DR) to reduce frequent modifications. The proposed method visualizes the design in three dimensions for its completeness and flawless representation. Based on the consumer recommendation, the lack of design flaws in the representation is identified, and multiple detections are presented. This is required to improve consumer satisfaction and the multi-dimensional projection between flaws and complete attire products. The learning is trained using the fixable representation, and therefore, the previous unsuitable designs are repelled by different recommendations. This improves the design adaptability, recommendation ratio, and representation ratio. Besides, it reduces the recommendation time and flaws.

Keywords

• 3D
• Deep learning
• Fashion design
• Flaw detection
• Recommendation system