项目作者: dikshantsagar

项目描述 :
Repository for PAI-BPR a state of the art Fashion recommendation system capturing user personal preference and attribute interpretability
高级语言: Jupyter Notebook
项目地址: git://github.com/dikshantsagar/PAI-BPR.git
创建时间: 2020-06-07T12:48:02Z
项目社区:https://github.com/dikshantsagar/PAI-BPR

开源协议:

下载


PAI-BPR

State of the art fashion recommendation system capturing user preference and capability of attribute preference interpretation

PWC

  • The whole project was implemented on Google Colab, hence it is suggested to run it there itself!!

folder structure for Code files

  • Attribute_keras.ipynb - Attribute Representation Model Training
  • GetAttribute.ipynb - Getting Visual representations from the Attribute Model
  • GPBPR2.py - The Model File.
  • train.py - Training file for our state of the art model.
  • test.py - Testing our model
  • Testing_Model.pynb - Notebook visualising our SOTA model and its predictions and outputs
  • main.ipynb - To run train.py with the desired requirements
  • “Attribute_keras.ipynb” and “GetAttribute.ipynb” extract visual features from an image .
  • For the dataset please contact the original curators from this Paper

Use Citation

  1. @misc{sagar2020paibpr,
  2. title={PAI-BPR: Personalized Outfit Recommendation Scheme with Attribute-wise Interpretability},
  3. author={Dikshant Sagar and Jatin Garg and Prarthana Kansal and Sejal Bhalla and Rajiv Ratn Shah and Yi Yu},
  4. year={2020},
  5. eprint={2008.01780},
  6. archivePrefix={arXiv},
  7. primaryClass={cs.CV}
  8. }