Skip to content

tobefans/LSSED

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

16 Commits
 
 
 
 
 
 

Repository files navigation

LSSED

The dataset of the paper "LSSED: A Large-Scale Dataset and Benchmark for Speech Emotion Recognition".

Dataset

In view of copyright reasons, researchers who are interested in applying for this dataset, please read and sign the license (EULA.pdf) carefully and send it to Prof. Xing by email. To ensure that you are a staff member of a university or research institution, please:

  1. Use the official email address to apply
  2. Attach the official website (if any)

Pre-trained models

Our pre-trained models are released here (password: SCUTLAB626EMOTION). It contains three versions of PyResNet, with ResNet50, ResNet101 or ResNet152 as the backbone respectively.

The import and use of the pre-trained model are as follows:

model = torch.load('path_to_model.pth')
output = model(input)

These pre-trained models can be directly applied to the classification task of four kinds of emotions, including "Angry(0)", "Neutral(1)", "Happy(2)" and "Sad(3)". If the user needs to perform other emotion recognition or related speech downstream tasks, then fine-tuning is necessary. The user can replace the fully connected layer classifier of the last layer of the model called "fc".

model_ft = torch.load('path_to_model.pth')
num_fc_ftr = model_ft.fc.in_features
model_ft.fc = nn.Linear(num_fc_ftr, num_class)

Contact

Prof. Xing: xfxing@scut.edu.cn

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages