Arjun et al’s ViT with the AMIGOS Dataset

In this case, we introduce how to use TorchEEG to train a vision transformer proposed by Arjun et al. on the AMIGOS dataset for emotion classification.

import logging
import os
import random
import time

import numpy as np
import torch
from torch.utils.data.dataloader import DataLoader
from torcheeg import transforms
from torcheeg.datasets import AMIGOSDataset
from torcheeg.datasets.constants.emotion_recognition.amigos import \
    AMIGOS_CHANNEL_LOCATION_DICT
from torcheeg.model_selection import KFoldGroupbyTrial
from torcheeg.models import ArjunViT
from torcheeg.trainers import ClassificationTrainer

Pre-experiment Preparation to Ensure Reproducibility

Use the logging module to store output in a log file for easy reference while printing it to the screen.

os.makedirs('./tmp_out/examples_amigos_arjunvit/log', exist_ok=True)
logger = logging.getLogger('Arjun et al\'s ViT with the AMIGOS Dataset')
logger.setLevel(logging.DEBUG)
console_handler = logging.StreamHandler()
timeticks = time.strftime("%Y-%m-%d-%H-%M-%S", time.localtime())
file_handler = logging.FileHandler(
    os.path.join('./tmp_out/examples_amigos_arjunvit/log', f'{timeticks}.log'))
logger.addHandler(console_handler)
logger.addHandler(file_handler)

Set the random number seed in all modules to guarantee the same result when running again.

def seed_everything(seed):
    random.seed(seed)
    np.random.seed(seed)
    os.environ["PYTHONHASHSEED"] = str(seed)
    torch.manual_seed(seed)
    torch.cuda.manual_seed(seed)
    torch.backends.cudnn.deterministic = True
    torch.backends.cudnn.benchmark = False


seed_everything(42)

Customize Trainer

TorchEEG provides a large number of trainers to help complete the training of classification models, generative models and cross-domain methods. Here we choose the simplest classification trainer, inherit the trainer and overload the log function to save the log using our own defined method; other hook functions can also be overloaded to meet special needs.

class MyClassificationTrainer(ClassificationTrainer):
    def log(self, *args, **kwargs):
        if self.is_main:
            logger.info(*args, **kwargs)

Building Deep Learning Pipelines Using TorchEEG

Step 1: Initialize the Dataset

We use the AMIGOS dataset supported by TorchEEG. Here, we set an EEG sample to 1 second long and include 128 data points. The baseline signal is 5 seconds long, cut into five, and averaged as the baseline signal for the trial. In offline preprocessing, all EEG signals are debaselined and normalized, and the preprocessed EEG signals are stored in the local IO. In online processing, all EEG signals are converted into Tensors for input into neural networks.

dataset = AMIGOSDataset(io_path='./tmp_out/examples_amigos_arjunvit/amigos',
                        root_path='./tmp_in/data_preprocessed',
                        offline_transform=transforms.Compose([
                            transforms.BaselineRemoval(),
                            transforms.MeanStdNormalize()
                        ]),
                        online_transform=transforms.ToTensor(),
                        label_transform=transforms.Compose([
                            transforms.Select('valence'),
                            transforms.Binary(5.0)
                        ]),
                        chunk_size=128,
                        baseline_chunk_size=128,
                        num_baseline=5,
                        num_worker=4,
                        io_mode='pickle')

Warning

If you use TorchEEG under the Windows system and want to use multiple processes (such as in dataset or dataloader), you should check whether __name__ is __main__ to avoid errors caused by multiple import.

That is, under the Windows system, you need to:

if __name__ == "__main__":
    dataset = AMIGOSDataset(io_path='./tmp_out/examples_amigos_arjunvit/amigos',
                     root_path='./tmp_in/data_preprocessed',
                     offline_transform=transforms.Compose([
                         transforms.BaselineRemoval(),
                         transforms.MeanStdNormalize()
                     ]),
                     online_transform=transforms.ToTensor(),
                     label_transform=transforms.Compose([
                         transforms.Select('valence'),
                         transforms.Binary(5.0)
                     ]),
                     io_mode='pickle',
                     chunk_size=128,
                     baseline_chunk_size=128,
                     num_baseline=5,
                     num_worker=4)
    # the following codes

Note

LMDB may not be optimized for parts of Windows systems or storage devices. If you find that the data preprocessing speed is slow, you can consider setting io_mode to pickle, which is an alternative implemented by TorchEEG based on pickle.

Step 2: Divide the Training and Test samples in the Dataset

Here, the dataset is divided using 5-fold cross-validation. In the process of division, we group according to the trial index, and every trial takes 4 folds as training samples and 1 fold as test samples. Samples across trials are aggregated to obtain training set and test set.

k_fold = KFoldGroupbyTrial(
    n_splits=5, split_path='./tmp_out/examples_amigos_arjunvit/split')

Step 3: Define the Model and Start Training

We first use a loop to get the dataset in each cross-validation. In each cross-validation, we initialize the ArjunViT model, and define the hyperparameters. For example, each EEG sample contains 128 time points, we divide it into 4 patches, each patch contains 32 time points, etc.

We then initialize the trainer and set the hyperparameters in the trained model, such as the learning rate, the equipment used, etc. The fit method receives the training dataset and starts training the model. The test method receives a test dataset and reports the test results. The save_state_dict method can save the trained model.

for i, (train_dataset, val_dataset) in enumerate(k_fold.split(dataset)):
    # Initialize the model
    model = ArjunViT(chunk_size=128,
                     t_patch_size=32,
                     hid_channels=32,
                     num_electrodes=len(AMIGOS_CHANNEL_LOCATION_DICT),
                     num_classes=2)

    # Initialize the trainer and use the 0-th GPU for training, or set device_ids=[] to use CPU
    trainer = MyClassificationTrainer(model=model,
                                      lr=1e-4,
                                      weight_decay=1e-4,
                                      device_ids=[0])

    # Initialize several batches of training samples and test samples
    train_loader = DataLoader(train_dataset,
                              batch_size=256,
                              shuffle=True,
                              num_workers=4)
    val_loader = DataLoader(val_dataset,
                            batch_size=256,
                            shuffle=False,
                            num_workers=4)

    # Do 50 rounds of training
    trainer.fit(train_loader, val_loader, num_epochs=50)
    trainer.test(val_loader)
    trainer.save_state_dict(
        f'./tmp_out/examples_amigos_arjunvit/weight/{i}.pth')