RAMer: Reconstruction-based Adversarial Model for Multi-party Multi-model Multi-label Emotion Recognition

This is an anonymous repository for double-blind manuscript review.

TODO List

• Overview

  • The model framework.

• Requirements

  • List all dependencies and libraries required to run the code.

• Training

  • Provide instructions on how to train the model, including command examples and parameter explanations.

• Evaluation

  • Explain the evaluation process and provide examples of how to evaluate the model's performance.

Overview

The framework of RAMer. Given incomplete multi-modal inputs, RAMer first encodes each individual modality through an auxiliary task, then feeds the features into a reconstruction-based adversarial network to extract specificity and commonality. Finally, a stacked shuffle layer is employed to learn enhanced representations.

Requirements

To install requirements:

pip install -r requirements.txt

📋 Describe how to set up the environment, e.g. pip/conda/docker commands, download datasets, etc...

Training

To train the model(s) in the paper, run this command:

sh train.sh

📋 Describe the training details, including the full training procedure and appropriate hyperparameters.

Evaluation

To evaluate my model on CMU-MOSEI, M3ED and MEmoR, run:

inclued in train.sh

📋 Describe the evaluation process, and give commands that produce the results.

Models

You can download the checkpoint of models here:

• [model].

Contributing

📋 Pick a licence.