Welcome to SFTF5000

SFTF5000 is a unique dataset that contains 5,000+ audio files converted to Short-Time Fourier Transform (STFT) representations. This dataset is designed for researchers, data scientists, and AI enthusiasts working on signal processing, deep learning, and audio analysis.

This dataset is securely hosted on Amazon AWS S3 for fast and reliable access.

Click the buttons below to download the dataset directly from AWS S3.

Download SFTF5000 Dataset Download SFTF5000 (Normalized) Dataset

Mark Jerome De Chavez, March 2025