Audio deepdream: Optimizing raw audio with convolutional networks

The hallucinatory images of DeepDream [8] opened up the floodgates for a recent wave of artwork generated by neural networks. In this work, we take first steps to applying this to audio. We believe a key to solving this problem is training a deep neural network to perform a perception task on raw audio. Consequently, we have followed in the footsteps of Van den Oord et al [13] and trained a network to predict embeddings that were themselves the result of a collaborative filtering model. A key difference is that we learn features directly from the raw audio, which creates a chain of differentiable functions from raw audio to high level features. We then use gradient descent on the network to extract samples of” dreamed” audio. Examples are available at http://tiny. cc/78qqdy.