Giter Site home page Giter Site logo

ba_cnn's Introduction

An Attention-Based Architecture for Hierarchical Classification with CNNs

This directory contains an implementation of BA-CNN paper (link). In our experiments, we used CIFAR-10,CIFAR-100 and Fashion MNIST, and compared our proposed method (BA-CNN) results with B-CNN (Zhu and Bain, 2017), H-CNN (Zhang,2021), Add-net and Concat-net (Inoue,2020)

Requirements

For the experiments, CIFAR-10 and Fashion MNIST were imported from keras library. CIFAR-100 can be downloaded from https://www.cs.toronto.edu/~kriz/cifar.html

Files

/code/

  • BA_CNN_CIFAR_10.ipynb: Models B-CNN, H-CNN, Add-net, Concat-net and BA-CNN Base B and C, running on CIFAR-10 dataset.
  • BA_CNN_CIFAR_100.ipynb: Models B-CNN, H-CNN, Add-net, Concat-net and BA-CNN Base C, running on CIFAR-100 dataset.
  • BA_CNN_CIFAR_FASHION_MNIST.ipynb: Models B-CNN, H-CNN, Add-net, Concat-net and BA-CNN Base C, running on Fashion MNIST dataset.

/output/

  • html files from executed jupyter notebooks

References

B-CNN

@article{zhu2017b,
  title={B-CNN: branch convolutional neural network for hierarchical classification}, 
  author={Zhu, Xinqi and Bain, Michael},
  journal={arXiv preprint arXiv:1709.09890},
  year={2017}
}

Sample code on https://github.com/zhuxinqimac/B-CNN

H-CNN

@article{zhang2021hierarchical,
  title={Hierarchical bilinear convolutional neural network for image classification},
  author={Zhang, Xiang and Tang, Lei and Luo, Hangzai and Zhong, Sheng and Guan, Ziyu and Chen, Long and Zhao, Chao and Peng, Jinye and Fan, Jianping},
  journal={IET Computer Vision},
  volume={15},
  number={3},
  pages={197--207},
  year={2021},
  publisher={Wiley Online Library}
}

Sample code not available

B-CNN applied to Fashion MNIST

@article{seo2019hierarchical,
  title={Hierarchical convolutional neural networks for fashion image classification},
  author={Seo, Yian and Shin, Kyung-shik},
  journal={Expert systems with applications},
  volume={116},
  pages={328--339},
  year={2019},
  publisher={Elsevier}
}

Sample code on https://github.com/AdicherlaVenkataSai/HCNN

Add-net and Concat-net

@inproceedings{inoue2020semantic,
  title={Semantic Hierarchy-based Convolutional Neural Networks for Image Classification},
  author={Inoue, Matheus and Forster, Carlos Henrique and dos Santos, Antonio Carlos},
  booktitle={2020 International Joint Conference on Neural Networks (IJCNN)},
  pages={1--8},
  year={2020},
  organization={IEEE}
}

Sample code not available

ba_cnn's People

Contributors

ivanpizarroq avatar

Stargazers

 avatar

Watchers

 avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.