User Tools
**This is an old revision of the document!** ----
====== DeepDreamEffect ====== | **Authors **| Christian Dittmar | | **Affiliation** | International Audio Laboratories Erlangen | | **eMail** | [[christian.dittmar@audiolabs-erlangen.de]] | ===== What I did ===== I used Google's DeepDream processing as an audio effect. Therefore, I export music magnitude spectrogram as RGB channels of PNG images and apply to 'Gradient Ascent' with pre-trained networks to these images. Afterwards, I import these images again and resynthesize them using Griffin and Lim's method. {{ :overview.png?nolink&400 |}} Since the networks were trained on natural images, this makes no sense musically. However, it gives interesting results: ===== Example 1: Piano ===== {{ :shenua.wav |}} {{ :output_shenhua_layer3.wav |}} {{ :output_shenhua_layer5.wav |}} ===== Example 2: Ethno ===== {{ :olcay.wav |}} {{ :output_olcay_layer3.wav |}} ===== Example 3: Separated Breakbeat ===== {{ :amenbrotherbreaknorm_mix.wav |}} {{ :output_amen_layer3.wav |}}
