Audio & Speech Datasets

xTorch provides a rich collection of built-in dataset handlers for a wide range of audio and speech processing tasks, from recognition and classification to synthesis and source separation.

All audio datasets are located under the xt::datasets namespace and can be found in the <xtorch/datasets/audio_processing/> header directory.

General Usage

The workflow for using an audio dataset is similar to that of other domains. You typically instantiate the dataset class with the path to your data and an optional pipeline of audio-specific transformations.

#include <xtorch/xtorch.h>
 
int main() {
    // 1. Define audio-specific transformations (optional)
    // For example, creating a Mel Spectrogram from the raw audio waveform.
    auto transforms = std::make_unique<xt::transforms::Compose>(
        std::make_shared<xt::transforms::signal::MelSpectrogram>()
    );
 
    // 2. Instantiate a dataset for the Speech Commands task
    // The dataset will handle downloading and loading the data.
    auto dataset = xt::datasets::SpeechCommands(
        "./data",
        xt::datasets::DataMode::TRAIN,
        /*download=*/true,
        std::move(transforms)
    );
 
    std::cout << "Speech Commands dataset size: " << *dataset.size() << std::endl;
 
    // 3. Pass the dataset to a DataLoader
    xt::dataloaders::ExtendedDataLoader data_loader(dataset, 32, true);
 
    // The data loader is now ready for use in a training loop
    for (auto& batch : data_loader) {
        auto spectrograms = batch.first;
        auto labels = batch.second;
        // ... training step ...
    }
}

!!! info "Dataset Constructors" Most dataset constructors follow a standard pattern: DatasetName(const std::string& root, DataMode mode, bool download, TransformPtr transforms)

• root: The directory where the data is stored or will be downloaded.
• mode: DataMode::TRAIN, DataMode::TEST, or DataMode::VALIDATION.
• download: If true, the dataset will be downloaded if not found in the root directory.
• transforms: A unique_ptr to a transform pipeline to be applied to the data.

Available Datasets by Task

Audio Event Detection

Binary Speech Classification

Emotion Recognition

Environmental Sound Classification

Intent Classification

Music Genre Classification

Music Information Retrieval

Music Source Separation

Music Tagging

Sound Event Detection

Speaker Identification and Verification

Speech Command Recognition

Speech Recognition

Speech Separation

Speech Synthesis