GarmentIQ: Automated Garment Measurement for Fashion Retail | Dashboard

Progress & Tentative Schedule

Status: ✓ - completed, ⭘ - in progress, ⨯ - scheduled.

• Classification

  • ✓ Data scraping (click to see detail)
  • ✓ Data exploration
  • ✓ Train base models
  • ✓ Model fine tune and compare
  • ✓ Orgnaize code following the Python library standard
  • ✓ Connect to next steps - segmentation and measurement instruction

• Segmentation

  • ✓ Find and apply pre-trained model
  • ✓ Orgnaize code following the Python library standard
  • ✓ Connect to next step - landmark extraction

• Measurement instruction

  • ✓ Read measurement instruction
  • ✓ Transfer instruction to machine readable format
  • ✓ Connect to next step - landmark extraction

• Landmark extraction

  • ✓ Find and apply pre-trained model
  • ✓ Model fine tune and adjustment
  • ✓ Orgnaize code following the Python library standard

• Web interface demo

  • ✓ Application - tailor (the whole pipeline)
  • ✓ Garment measurement instruction generation
  • ✓ Garment image classification
  • ✓ Garment image segmentation
  • ✓ Garment landmark detection
  • ✓ Garment landmark adjustment

• Documentation

  • ✓ MagicBox
  • ✓ Python API

Authors

1. Name: Fangzhou Ma

Institution: Swiss Federal Institute of Technology Zurich

2. Name: Li Yuan

Institution: Swiss Federal Institute of Technology Zurich

3. Name: Xinrui Zhai

Institution: Swiss Federal Institute of Technology Zurich

Challenge Giver & Coach

1. Challenge Giver: Adrian Gonzalez-Sieira

Organization: INDITEX

2. Challenge Giver: Laura Rodriguez-Barreiro

Organization: INDITEX

3. Coach: Elliott Ash

Organization: Swiss Federal Institute of Technology Zurich

Data Science Lab Materials

🔒 Google Drive: Open in a new window

Report

Poster

Open in a new window

GitHub Repository