Datasets:

---

safe-challenge

/

safe-challenge-practice-dataset

like 0

Follow

SAFE Challenge 24

This pull request significantly improves the dataset card for the SAFE Challenge practice dataset by:

• Adding task_categories: ['audio-classification'] to the metadata, making the dataset discoverable by task.
• Adding relevant tags: ['audio-forensics', 'synthetic-speech-detection', 'challenge', 'text-to-speech'] and language: ['en'] for better discoverability and context.
• Including links to the official paper (Audio Forensics Evaluation (SAFE) Challenge), the project page (https://stresearch.github.io/SAFE/), and the GitHub repository (https://github.com/stresearch/SAFE).
• Adding a "Sample Usage" section with a Python code snippet to demonstrate how to load the dataset using the datasets library, clarifying the handling of undecoded audio.
• Expanding the dataset description for better clarity, drawing from the paper abstract and GitHub README.

These changes will enhance the dataset's discoverability and provide users with comprehensive information for its usage and context within the SAFE Challenge.