Falah/classification_arabic_dialects

Classification of Arabic Dialects Audio Dataset

This dataset contains audio samples of various Arabic dialects for the task of classification and recognition. The dataset aims to assist researchers and practitioners in developing models and systems for Arabic spoken language analysis and understanding.

Dataset Details

• Dataset Name: Classification of Arabic Dialects Audio Dataset
• Dataset URL: Falah/classification_arabic_dialects
• Dataset Size: 166,407,297 bytes
• Download Size: 158,117,904 bytes
• Splits:
  • Train: 130 examples

Class Labels and Mapping

The dataset consists of audio samples from the following Arabic dialects, along with their corresponding class labels:

• '0': Algeria
• '1': Egypt
• '2': Iraq
• '3': Jordan
• '4': Morocco
• '5': Saudi Arabia
• '6': Sudan
• '7': Syria
• '8': Tunisia
• '9': Yemen

Please refer to the dataset for the audio samples and their respective class labels.

Usage Example

To play and display an audio sample from the dataset, you can use the following code:

from IPython.display import Audio

country_names = ['Algeria', 'Egypt', 'Iraq', 'Jordan', 'Morocco', 'Saudi_Arabia', 'Sudan', 'Syria', 'Tunisia', 'Yemen']

index = 0  # Index of the audio example
label = dataset["train"][index]["label"]

country_name = country_names[int(label)]

audio_data = dataset["train"][index]["audio"]["array"]
sampling_rate = dataset["train"][index]["audio"]["sampling_rate"]

# Play audio
display(Audio(audio_data, rate=sampling_rate))

print("Class Label:", label)
print("Country Name:", country_name)

Make sure to replace index with the desired index of the audio example. This code will play the audio, display it, and print its associated class label and the matched country name from the country_names list.

Applications

The Classification of Arabic Dialects Audio Dataset can be utilized in various applications, including but not limited to:

• Arabic dialect classification
• Arabic spoken language recognition
• Speech analysis and understanding for Arabic dialects
• Acoustic modeling for Arabic dialects
• Cross-dialect speech processing and synthesis

Feel free to explore and leverage this dataset for your research and development tasks related to Arabic spoken language analysis and recognition.

License

The dataset is made available under the terms of the Creative Commons Attribution-ShareAlike 4.0 International (CC BY-SA 4.0) license.

Citation

If you use this dataset in your research or any other work, please consider citing it as

For more information or inquiries about the dataset, please contact the dataset author(s) mentioned in the citation.

@dataset{classification_arabic_dialects,
  author = {Falah.G.Salieh},
  title = {Classification of Arabic Dialects Audio Dataset},
  year = {2023},
  publisher = {Hugging Face},
  url = {https://huggingface.co/datasets/Falah/classification_arabic_dialects},
}