This experiment explores creating ambient sounds through machine learning process. The 24 unique forest sounds are generated by feature extraction from Magenta’s NSynth model.
