Skip to content

Implementation of the paper Adding Attentiveness to the Neurons in Recurrent Neural Network for action classification

Notifications You must be signed in to change notification settings

rvk7895/EleAtt-RNN-for-Action-Classification

Repository files navigation

SMAI_S22_48

Team Name : Statistically Significant

Members

  • Ritvik Aryan Kalra Github: rvk7895 Roll number - 2019115002
  • Lalit Kumar Digalla

This repository consists of implementation of the paper Adding Attentiveness to the Neurons in Recurrent Neural Networks

Datasets Used

  • Joint-annotated Human Motion Data Base Link
  • UCF101 - Action Recognition Data Set Link

Links

Presentation - Canva

Models and Dataset - OneDrive

Remote Repository - Github

Files

att_gru_jhmdb.ipynb - consists of the Vanilla and with EleAtt gate implementation of GRU on the JHMDB dataset
att_gru_ucf101.ipynb - consists of the Vanilla and with EleAtt gate implementation of GRU on the UCF101 dataset
att_lstm_jhmdb.ipynb - consists of the Vanilla and with EleAtt gate implementation of LSTM on the JHMDB dataset
att_lstm_ucf101.ipynb - consists of the Vanilla and with EleAtt gate implementation of LSTM on the UCF101 dataset\

TODOs

  • Implementations of Standard GRU
  • Implementaiton of GRU with attention layer
  • Implementation of Standard LSTM
  • Implementation of LSTM with attention layer
  • Implementation of Vanilla RNN
  • Implementation of Vanilla RNN with attention layer

About

Implementation of the paper Adding Attentiveness to the Neurons in Recurrent Neural Network for action classification

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published