DeepCodiNet
Introduction

- DeepCodiNet is a project evaluating people’s fashion based on deep learning method.
- The project is a part of SKKU’s Open Source Software Practice, and the project progress data is well visualized in https://skku-oss-10.github.io/
Dataset
We use following databases as our data set:
- Training set: 2016~2019 MUSINSA styling dataset (available at: https://store.musinsa.com/app/styles/lists?sex=&use_yn_360=&brand=&model=&max_rt=2019&min_rt=2010 &year_date=2016&month_date=&display_cnt=60&list_kind=small&sort=rt&page=)
- Test set: Clothing-Co-Parsing dataset (available at: https://github.com/bearpaw/clothing-co-parsing/)
Model
We implemented the detection model, which has been trained with DeepFashion2 dataset.
Due to the size of the trained model, we couldn’t upload the whole file.
Download is available at: https://drive.google.com/open?id=1Nrjqs3PdW5FPlhWEIAWyRTW8LUtCYutC