ibab/tensorflow-wavenet

This is a TensorFlow implementation of the WaveNet generative neural network architecture for audio generation. The WaveNet architecture directly generates a raw audio waveform, and shows excellent results in TTS and general audio generation (see the DeepMind blog post and paper for examples). The network is a model of the conditional probability to generate the next sample in the audio waveform, given all previous samples and possibly additional parameters. It is constructed from a stack of causal dilated layers, each of which is a dilated convolution (convolution with holes), which only accesses the current and past audio samples. The network is implemented in the file wavenet.py.