NExTplusplus / FashionKE

FashionKE is an effective implementation of an automatic harvesting model for fashion concept prediction and a comprehensive analysis tool for fashion knowledge extraction.

FashionKE

Code of paper: "Who, Where, and What to Wear? Extracting Fashion Knowledge from Social Media"

Requirements

python>=3  
pytorch==0.4.1  
torchvision==6.2.0 
pyyaml==5.1.2
json==2.0.9  
numpy==1.18.1 
pickle
tensorboard_logger==0.1.0
scipy==1.18.1 

Train

You can modify the config file in config.yaml to change the hyper-parameters.

python train.py

Log and Results

The default tensorboard log files are saved at ./checkpoint, where you can start a tensorboard to see the training curve.
The default trained model parameters are saved at ./params.
The default prediction results are saved at ./result.

Acknowledgement

This research is supported by the National Research Foundation, Singapore under its International Research Centres in Singapore Funding Initiative. Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not reflect the views of National Research Foundation, Singapore.