인공지능 readme

1. 인공지능 모델 소개

MMfashion에서 제공하는 다음과 같은 사전 훈련된 모델을 사용함.

1. Category and Attribute Prediction : 의류 카테고리/속성 분류 모델
2. In-Shop Clothes Retrieval : 의류 유사도 분석 모델

2. Setup

• MMfashion Github 주소 : https://github.com/open-mmlab/mmfashion

git clone --recursive https://github.com/open-mmlab/mmfashion.git
cd mmfashion
python setup.py install

3. Backbone

참고: https://github.com/open-mmlab/mmfashion/blob/master/docs/MODEL_ZOO.md

Category and Attribute Prediction

Backbone	Pooling	Loss	Top-5 Cate. Recall	Top-5 Attr. Recall.	Download (Google Drive)
VGG-16	Global Pooling	Cross-Entropy	35.91	25.44	model

In-Shop Clothes Retrieval

Backbone	Pooling	Loss	Top-5 Acc.	Download (Google Drive)
VGG-16	Global Pooling	Cross-Entropy	38.76	model

4. MMFashion 데이터 셋

• Readme : https://github.com/open-mmlab/mmfashion/blob/master/docs/DATA_PREPARATION.md
• 데이터 다운로드 : http://mmlab.ie.cuhk.edu.hk/projects/DeepFashion/AttributePrediction.html

5. MMFashion 사전 훈련된 모델 테스트

참고: https://github.com/open-mmlab/mmfashion/blob/master/docs/GETTING_STARTED.md