Sample-level Deep Convolutional Neural Networks for Music Auto-tagging Using Raw Waveforms
Paper • 1703.01789 • Published
This is a SampleCNN model which is used to distinguish between voice and music, that is Speech Music Discrimation.
For details on SampleCNNs see Sample-level Deep Convolutional Neural Networks for Music Auto-tagging Using Raw Waveforms by Jongpil Lee, Jiyoung Park, Keunhyoung Luke Kim, Juhan Nam.
This model was built using code at codeberg.org/ook/mysterytrain/training, see that repository for more details.
The original code this is based on is from github.com/pRecSys/samplecnn-speech-detection. The only additions this makes to the original are some code updates/cleanups so I could train on a Macbook with more recent python, librosa etc and some additional voice samples added to the training data set which I prepared manually.
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support