What happens if you apply the same deep dream technique which produces "dream" visuals but to media streams such as audio files?

Does changing image functions into audio and enhancing the logic would work, or will it no longer work/doesn't make any sense?

My goal is to create "dream" like audio based on the two samples.


As far as I can see, there's no reason why you couldn't (for example) take the convolutional inputs to deepdream from adjacent sample points, rather than adjacent spatial positions, as is the case with image input.

Given the 'self similar' nature of deep dream images, listening to this fractal granular synthesis technique might be of interest/inspiration.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.