This repository implements our IEEE Access (Q1 Journal) research:
Automatic Classification and Color Changing of Saree Components Using Deep Learning Techniques.
We present a deep learning pipeline for saree component classification. Our system enables retailers and designers to automatically classify saree parts and visualize color/pattern changes without manual editing. The proposed approach achieves 93.01% classification accuracy, significantly outperforming prior methods for textile image analysis and boosting the speed of design workflows.
- Highlights:
- Automatic segmentation & classification of saree components
- SaaS-ready pipeline for color & pattern transformation
- State-of-the-art accuracy using Mask R-CNN & VGG-16
- Designed for textile retailers, designers, and e-commerce applications
For further details, see our IEEE Access paper (2024).