File size: 515 Bytes
6bea901 071f1eb 6bea901 071f1eb 6bea901 071f1eb 6bea901 071f1eb 6bea901 071f1eb | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 | ---
license: mit
---
A deep learning model for audio classification, designed to detect AI-generated music.
Model architecture : EfficientNetB0 (Transfer Learning)
Framework: TensorFlow / Keras
Task: Binary Classification (AI Music vs. Real Music)
Input: Mel Spectrograms (224x224) generated from audio files.
Training Data: Trained on a balanced dataset of thousands of
modern songs from sources like the Free Music Archive, NCS,
and the Suno AI dataset on Kaggle, all standardized to a
consistent bitrate. |